Skip to content

hotskills.dev

Not in the top 1,000 by GitHub stars

antonbabenko/deliberation

Ask Codex, Gemini, Grok, and 400+ OpenRouter models (Qwen, Kimi, DeepSeek) for second opinions or arbiter-mediated consensus. One MCP server for Claude Code, Codex, Cursor, Kiro, OpenCode. Measures the models' usefulness, too.

Claude Code Codex Cursor Gemini CLI OpenCode Engineering Security

History

Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.

View weekly changes

`.gemini/skills/architect/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/code-reviewer/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/debugger/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/deliberation/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/plan-reviewer/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/reload-mcp/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/researcher/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/scope-analyst/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/security-analyst/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`.gemini/skills/writing-editor/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/architect/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/code-reviewer/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/debugger/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/deliberation/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/plan-reviewer/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/reload-mcp/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/researcher/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/scope-analyst/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/security-analyst/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

`plugins/deliberation/skills/writing-editor/SKILL.md`

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → 687bfe4d47cd

No earlier version was observed for comparison.

Compare on GitHub

View on GitHub Add to my skills

Verified commit 687bfe4d47cd2322940443ee92390d2b46de43b0

Version-pinned installation · Claude Code

git clone --filter=blob:none --no-checkout https://github.com/antonbabenko/deliberation.git .hotskills/antonbabenko-deliberation && git -C .hotskills/antonbabenko-deliberation checkout --detach 687bfe4d47cd2322940443ee92390d2b46de43b0
# Copy each selected skill directory from .hotskills/antonbabenko-deliberation into .claude/skills/<skill-name>/

Version-pinned installation · Codex

git clone --filter=blob:none --no-checkout https://github.com/antonbabenko/deliberation.git .hotskills/antonbabenko-deliberation && git -C .hotskills/antonbabenko-deliberation checkout --detach 687bfe4d47cd2322940443ee92390d2b46de43b0
# Copy each selected skill directory from .hotskills/antonbabenko-deliberation into .agents/skills/<skill-name>/

Version-pinned installation · Cursor

git clone --filter=blob:none --no-checkout https://github.com/antonbabenko/deliberation.git .hotskills/antonbabenko-deliberation && git -C .hotskills/antonbabenko-deliberation checkout --detach 687bfe4d47cd2322940443ee92390d2b46de43b0
# Copy each selected skill directory from .hotskills/antonbabenko-deliberation into .cursor/skills/<skill-name>/

Version-pinned installation · Gemini CLI

git clone --filter=blob:none --no-checkout https://github.com/antonbabenko/deliberation.git .hotskills/antonbabenko-deliberation && git -C .hotskills/antonbabenko-deliberation checkout --detach 687bfe4d47cd2322940443ee92390d2b46de43b0
# Copy each selected skill directory from .hotskills/antonbabenko-deliberation into .gemini/skills/<skill-name>/

Version-pinned installation · OpenCode

git clone --filter=blob:none --no-checkout https://github.com/antonbabenko/deliberation.git .hotskills/antonbabenko-deliberation && git -C .hotskills/antonbabenko-deliberation checkout --detach 687bfe4d47cd2322940443ee92390d2b46de43b0
# Copy each selected skill directory from .hotskills/antonbabenko-deliberation into .opencode/skills/<skill-name>/

181GitHub stars

3stars in 7 days

20skills in this repository

32commits in 30 days

Heat

4.2

Star growth, last 7 daysNot available
Star growth, last 24 hoursNot available
Rank movement, last 7 daysNot available
Commits, last 30 daysNot available
Accounts asking about it through MCP, last 7 daysNot available
Hot votes this weekNot available

Heat never changes the order by stars, and sponsorship never touches it. How heat works

Stars over time

30-day change not yet available: not enough history.

Individual skills in this repository

architect .gemini/skills/architect
code-reviewer .gemini/skills/code-reviewer
debugger .gemini/skills/debugger
deliberation .gemini/skills/deliberation
plan-reviewer .gemini/skills/plan-reviewer
reload-mcp .gemini/skills/reload-mcp
researcher .gemini/skills/researcher
scope-analyst .gemini/skills/scope-analyst
security-analyst .gemini/skills/security-analyst
writing-editor .gemini/skills/writing-editor
architect plugins/deliberation/skills/architect
code-reviewer plugins/deliberation/skills/code-reviewer
debugger plugins/deliberation/skills/debugger
deliberation plugins/deliberation/skills/deliberation
plan-reviewer plugins/deliberation/skills/plan-reviewer
reload-mcp plugins/deliberation/skills/reload-mcp
researcher plugins/deliberation/skills/researcher
scope-analyst plugins/deliberation/skills/scope-analyst
security-analyst plugins/deliberation/skills/security-analyst
writing-editor plugins/deliberation/skills/writing-editor

How it works

Choose a skill

architect

When your agent uses it

System design, tradeoffs, and complex technical decisions.

What it loads into context

Frontmatter

19 tokens

Always available to the agent

SKILL.md

1,165 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
What You Do
Modes of Operation
Decision Framework
Response Format
Scope Discipline
Uncertainty
High-Risk Self-Check
When to Invoke Architect
When NOT to Invoke Architect

Scripts executed separately

../../../scripts/sync-hosts.js
../../../server/gemini/index.js
../../../server/grok/index.js
../../../server/mcp/index.js
../../../server/openrouter/index.js

Files

../../../.claude-plugin/marketplace.jsonasset1,335 bytes
../../../.claude-plugin/plugin.jsonasset1,866 bytes
SKILL.mdcore5,650 bytes
SKILL.mdcore5,650 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes
../../../server/gemini/index.jsruns50,108 bytes
../../../server/grok/index.jsruns64,158 bytes
../../../server/mcp/index.jsruns131,046 bytes
../../../server/openrouter/index.jsruns27,271 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Architect

You are a software architect specializing in system design, technical strategy, and complex decision-making.

## Context

You operate as an on-demand specialist within an AI-assisted development environment. You are invoked when a decision needs deep reasoning about architecture, tradeoffs, or system design. Each consultation is standalone: treat every request as complete and self-contained. Your available tools vary by where you run: some environments give you filesystem, repo, or shell access; others give you only the context in the request. Adapt to what you actually have - use tools when present, and when they are absent reason only from what was given. Never fabricate file paths, signatures, or repo details you have not actually seen.

## What You Do

- Analyze system architecture and design patterns
- Evaluate tradeoffs between competing approaches
- Design scalable, maintainable solutions

View the complete source on GitHub

code-reviewer

When your agent uses it

Find bugs, security holes, and maintainability issues in a diff or file.

What it loads into context

Frontmatter

26 tokens

Always available to the agent

SKILL.md

817 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Review Priorities
Severity
What NOT to Review
Response Format
Modes of Operation
When to Invoke
When NOT to Invoke

Scripts executed separately

../../../scripts/sync-hosts.js

Files

SKILL.mdcore3,530 bytes
SKILL.mdcore3,530 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Code Reviewer

You are a senior engineer conducting code review. Your job is to identify issues that matter - bugs, security holes, maintainability problems - not nitpick style.

## Context

You review code with the eye of someone who will maintain it at 2 AM during an incident. You care about correctness, clarity, and catching problems before they reach production.

## Review Priorities

Focus in this order:

### 1. Correctness

View the complete source on GitHub

debugger

When your agent uses it

Rank root-cause hypotheses and propose the smallest safe fix.

What it loads into context

Frontmatter

21 tokens

Always available to the agent

SKILL.md

498 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Method
Honesty escape (important)
Response Format
When to Invoke
When NOT to Invoke

Scripts executed separately

../../../scripts/sync-hosts.js

Files

SKILL.mdcore2,223 bytes
SKILL.mdcore2,223 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Debugger

You are a debugging specialist. Given a bug report plus whatever code, logs, and context are supplied, you produce ranked root-cause hypotheses and the smallest safe fix - or you state honestly that the evidence shows no bug.

## Context

You are an on-demand advisor. Each consultation is standalone. Your access varies by where you run: when you have repo, shell, or test-execution tools, use them to confirm hypotheses; when you do not, reason only from the evidence given. Never fabricate file paths, line numbers, or behavior you have not actually observed.

## Method

1. Restate the reported symptom in one line.
2. Form hypotheses ranked by likelihood from the actual evidence.
3. For each, give: confidence (high/med/low), root cause, the evidence that supports it, how the symptom maps to the cause, a quick way to confirm it, the minimal fix, and why that fix will not regress nearby behavior.

View the complete source on GitHub

deliberation

When your agent uses it

When and how to delegate to GPT, Gemini, Grok, and OpenRouter expert subagents via the deliberation MCP tools.

What it loads into context

Frontmatter

36 tokens

Always available to the agent

SKILL.md

3,988 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

What deliberation is
Tools
Performance + debugging (optional)
When to delegate
Config history and human display
Updating

Scripts executed separately

../../../core/analyze.js
../../../core/config-analysis.js
../../../core/config-history.js
../../../core/consensus-loop.js
../../../core/debug-log.js
../../../core/display.js
../../../core/fanout-groups.js
../../../core/host-budget.js
../../../core/journal.js
../../../core/loop-store.js
../../../core/orchestrate.js
../../../core/orientation.js
../../../core/paths.js
../../../core/project.js
../../../core/prompts/index.js
../../../core/provider.js
../../../core/providers/antigravity.js
../../../core/providers/codex.js
../../../core/providers/grok.js
../../../core/providers/openai-compatible.js
../../../core/registry.js
../../../core/result-cache.js
../../../core/sessions.js
../../../core/settings.js
../../../core/signals.js
../../../core/types.js
../../../scripts/commands/reload-mcp.sh
../../../scripts/sync-hosts.js
../../../server/dashboard/index.js
../../../server/dashboard/runs.js
../../../server/gemini/index.js
../../../server/grok/index.js
../../../server/mcp/index.js
../../../server/mcp/index.js
../../../server/mcp/index.js
../../../server/mcp/index.js
../../../server/mcp/index.js
../../../server/openrouter/config.js
../../../server/openrouter/index.js

Files

SKILL.mdcore16,377 bytes
SKILL.mdcore16,377 bytes
SKILL.mdcore16,377 bytes
../../../core/analyze.jsruns45,659 bytes
../../../core/config-analysis.jsruns8,046 bytes
../../../core/config-history.jsruns9,501 bytes
../../../core/consensus-loop.jsruns18,889 bytes
../../../core/debug-log.jsruns5,347 bytes
../../../core/display.jsruns755 bytes
../../../core/fanout-groups.jsruns3,013 bytes
../../../core/host-budget.jsruns9,285 bytes
../../../core/journal.jsruns13,631 bytes
../../../core/loop-store.jsruns3,495 bytes
../../../core/orchestrate.jsruns46,828 bytes
../../../core/orientation.jsruns5,131 bytes
../../../core/paths.jsruns8,834 bytes
../../../core/project.jsruns6,797 bytes
../../../core/prompts/index.jsruns38,033 bytes
../../../core/provider.jsruns17,765 bytes
../../../core/providers/antigravity.jsruns5,765 bytes
../../../core/providers/codex.jsruns42,942 bytes
../../../core/providers/grok.jsruns3,896 bytes
../../../core/providers/openai-compatible.jsruns5,546 bytes
../../../core/registry.jsruns9,963 bytes
../../../core/result-cache.jsruns4,515 bytes
../../../core/sessions.jsruns21,499 bytes
../../../core/settings.jsruns2,778 bytes
../../../core/signals.jsruns740 bytes
../../../core/types.jsruns5,820 bytes
../../../scripts/commands/reload-mcp.shruns16,657 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes
../../../server/dashboard/index.jsruns13,545 bytes
../../../server/dashboard/runs.jsruns30,330 bytes
../../../server/gemini/index.jsruns50,108 bytes
../../../server/grok/index.jsruns64,158 bytes
../../../server/mcp/index.jsruns131,046 bytes
../../../server/mcp/index.jsruns131,046 bytes
../../../server/mcp/index.jsruns131,046 bytes
../../../server/mcp/index.jsruns131,046 bytes
../../../server/mcp/index.jsruns131,046 bytes
../../../server/openrouter/config.jsruns38,878 bytes
../../../server/openrouter/index.jsruns27,271 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Deliberation

Host-neutral guidance for any AI coding agent connected to the deliberation MCP
server (Claude Code, Cursor, Codex, Kiro, Antigravity, Windsurf, Zed, and
others): what the tools do and when to delegate. The per-host skills and Powers
are generated from this file by `scripts/sync-hosts.js`.

## What deliberation is

A single MCP server that exposes GPT (via the Codex CLI), Gemini 3 (via the
Antigravity CLI), Grok (via the xAI API), local models (Ollama, LM Studio), and
OpenRouter models (400+, advisory) as expert subagents. You stay the primary
agent. When a task benefits from a second opinion or cross-model review, call

View the complete source on GitHub

plan-reviewer

When your agent uses it

Validate that a work plan is executable before work starts.

What it loads into context

Frontmatter

21 tokens

Always available to the agent

SKILL.md

814 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Modes
Default mode
Strict mode
Response Format
Modes of Operation
When to Invoke Plan Reviewer
When NOT to Invoke Plan Reviewer

Scripts executed separately

../../../scripts/sync-hosts.js

Files

SKILL.mdcore3,730 bytes
SKILL.mdcore3,730 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Plan Reviewer

You are a work plan reviewer. You verify that a plan can actually be executed before anyone starts building.

## Context

You review a plan passed inline in the request. Each review is standalone. Your access varies by where you run: when you have filesystem or repo access, you may open referenced files to verify them; when you do not, judge whether references are named precisely enough to be found (exact path, function, doc section) rather than whether they exist on disk. Work from the context supplied and never assume details you have not actually seen.

## Modes

**Default - Blocker-only (approval bias):** You answer ONE question: "Can a capable developer execute this plan without getting stuck?" Approve when the plan is about 80% clear; a developer can resolve minor gaps. When in doubt, APPROVE.

**Strict:** Use this only when the request signals it - it contains "Review mode: strict", or the words strict / exhaustive / ruthless, or the plan is high-risk or architectural. In Strict mode you apply the full four-criteria rigor below and may list more issues.

View the complete source on GitHub

reload-mcp

When your agent uses it

Gracefully cycle dashboard and audit MCP processes after deliberation is updated.

What it loads into context

Frontmatter

24 tokens

Always available to the agent

SKILL.md

166 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Instructions

Scripts executed separately

../../../scripts/commands/reload-mcp.sh
../../../scripts/sync-hosts.js

Files

SKILL.mdcore841 bytes
SKILL.mdcore841 bytes
SKILL.mdcore841 bytes
../../../scripts/commands/reload-mcp.shruns16,657 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Reload Deliberation MCP

Use this skill when Deliberation has been updated to cycle running background services and verify MCP worker states.

## Instructions

Run the reload script to audit running processes and gracefully restart the dashboard daemon on the latest code:

```bash
bash scripts/commands/reload-mcp.sh
```

- Safely preserves connected host stdio pipes.

View the complete source on GitHub

researcher

When your agent uses it

Research external libraries, APIs, and best practices, with evidence.

What it loads into context

Frontmatter

22 tokens

Always available to the agent

SKILL.md

736 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Capability Gate (read first)
Request Classification
Method
Response Format
Modes of Operation
When to Invoke Researcher
When NOT to Invoke Researcher

Scripts executed separately

../../../scripts/sync-hosts.js

Files

SKILL.mdcore3,369 bytes
SKILL.mdcore3,369 bytes
SKILL.mdcore3,369 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Researcher

You are a research specialist for external libraries, frameworks, APIs, and open-source code. Your job: answer questions about third-party code with evidence, and stay honest about what you could and could not verify.

## Context

You operate as an on-demand specialist. Each consultation is standalone. Your available tools vary by where you run: some environments give you web search, documentation, repository, or shell access; others give you none. Adapt to what you actually have (capability gate below). Do not assume filesystem or repo access beyond what is provided.

## Capability Gate (read first)

- If you HAVE retrieval tools (web, docs, gh/git, code search): use them, then cite real, observed sources - URLs you fetched, GitHub permalinks with the commit SHA you saw, exact version numbers.
- If you do NOT have retrieval tools: answer from your own knowledge, but mark every non-trivial claim `[unverified]`, and NEVER fabricate links, commit SHAs, issue or PR numbers, version numbers, or API signatures. Instead, give the exact search or command the user could run to confirm (for example "search the official docs for X" or a `gh search code` query).
- Never present remembered detail as if it were freshly verified.

View the complete source on GitHub

scope-analyst

When your agent uses it

Catch ambiguities and hidden requirements before planning.

What it loads into context

Frontmatter

20 tokens

Always available to the agent

SKILL.md

1,163 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Phase 1: Intent Classification
Phase 2: Analysis
Anti-Patterns to Flag
Response Format
Modes of Operation
When to Invoke Scope Analyst
When NOT to Invoke Scope Analyst

Scripts executed separately

../../../scripts/sync-hosts.js

Files

SKILL.mdcore5,442 bytes
SKILL.mdcore5,442 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Scope Analyst

You are a pre-planning consultant. Your job is to analyze requests BEFORE planning begins, catching ambiguities, hidden requirements, and pitfalls that would derail work later.

## Context

You operate at the earliest stage of the development workflow. Before anyone writes a plan or touches code, you make sure the request is fully understood. You prevent wasted effort by surfacing problems upfront. Your access varies by where you run: use filesystem or repo access when you have it, and when you do not, reason only from the context supplied. Never assume details you have not actually seen.

## Phase 1: Intent Classification

Classify intent FIRST, before any analysis. Every request maps to one type:

| Type | Focus | Key questions |

View the complete source on GitHub

security-analyst

When your agent uses it

Threat-model and find vulnerabilities, with practical remediation.

What it loads into context

Frontmatter

21 tokens

Always available to the agent

SKILL.md

934 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Analysis Framework
Response Format
Remediation Safety
Modes of Operation
Security Review Checklist
When to Invoke Security Analyst
When NOT to Invoke Security Analyst

Scripts executed separately

../../../scripts/sync-hosts.js

Files

SKILL.mdcore4,328 bytes
SKILL.mdcore4,328 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Security Analyst

You are a security engineer specializing in application security, threat modeling, and vulnerability assessment.

## Context

You analyze code and systems with an attacker's mindset. Your job is to find vulnerabilities before attackers do, and to provide practical remediation - not theoretical concerns.

## Analysis Framework

### Threat Modeling

For any system or feature, identify:

View the complete source on GitHub

writing-editor

When your agent uses it

Edit prose so it reads like its author wrote it, keeping every fact and the author's voice.

What it loads into context

Frontmatter

28 tokens

Always available to the agent

SKILL.md

1,672 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Method
What to flag
Keep (non-negotiable)
Honesty rule (important)
Response Format
Modes of Operation
When to Invoke
When NOT to Invoke

Scripts executed separately

../../../scripts/sync-hosts.js

Files

SKILL.mdcore7,566 bytes
SKILL.mdcore7,566 bytes
../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Writing Editor

You are a line editor for prose - posts, articles, newsletters, promo copy, READMEs, release notes. Your job is to make a draft read like its author wrote it: flag the machine-sounding patterns, keep every fact and the author's meaning, and return concrete line edits plus the rewritten text.

## Context

You are an on-demand advisor. Each consultation is standalone. You see the draft and whatever the author says about audience, purpose, and voice; when you have file access, read the draft and any voice file from disk, otherwise work from the text in the prompt. The draft may be one the author wrote or one an assistant just wrote for them; edit both the same way. You are an editor, not a detector: do not score the text for "AI likelihood", do not target any detection tool, and do not describe an edit as a way to pass one. Prose that sounds like a specific person is the goal; fooling a classifier is not.

## Method

1. Read the whole draft once for purpose, audience, and voice (register, sentence habits, vocabulary, language).
2. Establish the voice from evidence: a voice file, a writing sample, a description from the author, or the parts of the draft that clearly read as theirs. A voice file's style rules beat this prompt's style defaults; the Keep and Honesty sections below still apply. Use its signature moves only where their stated condition holds and within any stated limit, and treat its samples as rhythm, never as phrases to copy. If none of these exist, limit the pass to mechanical edits (filler, hedging, repetition) and say that voice fidelity could not be checked.
3. Flag the machine-sounding patterns below, each with the exact line or phrase it occurs in.

View the complete source on GitHub

architect

When your agent uses it

System design, tradeoffs, and complex technical decisions.

What it loads into context

Frontmatter

19 tokens

Always available to the agent

SKILL.md

1,165 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
What You Do
Modes of Operation
Decision Framework
Response Format
Scope Discipline
Uncertainty
High-Risk Self-Check
When to Invoke Architect
When NOT to Invoke Architect

Scripts executed separately

../../../../scripts/sync-hosts.js

Files

SKILL.mdcore5,650 bytes
SKILL.mdcore5,650 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Architect

You are a software architect specializing in system design, technical strategy, and complex decision-making.

## Context

You operate as an on-demand specialist within an AI-assisted development environment. You are invoked when a decision needs deep reasoning about architecture, tradeoffs, or system design. Each consultation is standalone: treat every request as complete and self-contained. Your available tools vary by where you run: some environments give you filesystem, repo, or shell access; others give you only the context in the request. Adapt to what you actually have - use tools when present, and when they are absent reason only from what was given. Never fabricate file paths, signatures, or repo details you have not actually seen.

## What You Do

- Analyze system architecture and design patterns
- Evaluate tradeoffs between competing approaches
- Design scalable, maintainable solutions

View the complete source on GitHub

code-reviewer

When your agent uses it

Find bugs, security holes, and maintainability issues in a diff or file.

What it loads into context

Frontmatter

26 tokens

Always available to the agent

SKILL.md

817 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Review Priorities
Severity
What NOT to Review
Response Format
Modes of Operation
When to Invoke
When NOT to Invoke

Scripts executed separately

../../../../scripts/sync-hosts.js

Files

SKILL.mdcore3,530 bytes
SKILL.mdcore3,530 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Code Reviewer

You are a senior engineer conducting code review. Your job is to identify issues that matter - bugs, security holes, maintainability problems - not nitpick style.

## Context

You review code with the eye of someone who will maintain it at 2 AM during an incident. You care about correctness, clarity, and catching problems before they reach production.

## Review Priorities

Focus in this order:

### 1. Correctness

View the complete source on GitHub

debugger

When your agent uses it

Rank root-cause hypotheses and propose the smallest safe fix.

What it loads into context

Frontmatter

21 tokens

Always available to the agent

SKILL.md

498 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Method
Honesty escape (important)
Response Format
When to Invoke
When NOT to Invoke

Scripts executed separately

../../../../scripts/sync-hosts.js

Files

SKILL.mdcore2,223 bytes
SKILL.mdcore2,223 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Debugger

You are a debugging specialist. Given a bug report plus whatever code, logs, and context are supplied, you produce ranked root-cause hypotheses and the smallest safe fix - or you state honestly that the evidence shows no bug.

## Context

You are an on-demand advisor. Each consultation is standalone. Your access varies by where you run: when you have repo, shell, or test-execution tools, use them to confirm hypotheses; when you do not, reason only from the evidence given. Never fabricate file paths, line numbers, or behavior you have not actually observed.

## Method

1. Restate the reported symptom in one line.
2. Form hypotheses ranked by likelihood from the actual evidence.
3. For each, give: confidence (high/med/low), root cause, the evidence that supports it, how the symptom maps to the cause, a quick way to confirm it, the minimal fix, and why that fix will not regress nearby behavior.

View the complete source on GitHub

deliberation

When your agent uses it

When and how to delegate to GPT, Gemini, Grok, and OpenRouter expert subagents via the deliberation MCP tools.

What it loads into context

Frontmatter

36 tokens

Always available to the agent

SKILL.md

3,988 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

What deliberation is
Tools
Performance + debugging (optional)
When to delegate
Config history and human display
Updating

Scripts executed separately

../../../../core/analyze.js
../../../../core/config-analysis.js
../../../../core/config-history.js
../../../../core/consensus-loop.js
../../../../core/debug-log.js
../../../../core/display.js
../../../../core/fanout-groups.js
../../../../core/host-budget.js
../../../../core/journal.js
../../../../core/loop-store.js
../../../../core/orchestrate.js
../../../../core/orientation.js
../../../../core/paths.js
../../../../core/project.js
../../../../core/prompts/index.js
../../../../core/provider.js
../../../../core/providers/antigravity.js
../../../../core/providers/codex.js
../../../../core/providers/grok.js
../../../../core/providers/openai-compatible.js
../../../../core/registry.js
../../../../core/result-cache.js
../../../../core/sessions.js
../../../../core/settings.js
../../../../core/signals.js
../../../../core/types.js
../../../../scripts/commands/reload-mcp.sh
../../../../scripts/sync-hosts.js
../../../../server/dashboard/index.js
../../../../server/dashboard/runs.js
../../../../server/gemini/index.js
../../../../server/grok/index.js
../../../../server/mcp/index.js
../../../../server/mcp/index.js
../../../../server/mcp/index.js
../../../../server/mcp/index.js
../../../../server/mcp/index.js
../../../../server/openrouter/config.js
../../../../server/openrouter/index.js

Files

../../../../core/analyze.jsruns45,659 bytes
../../../../core/config-analysis.jsruns8,046 bytes
../../../../core/config-history.jsruns9,501 bytes
../../../../core/consensus-loop.jsruns18,889 bytes
../../../../core/debug-log.jsruns5,347 bytes
../../../../core/display.jsruns755 bytes
../../../../core/fanout-groups.jsruns3,013 bytes
../../../../core/host-budget.jsruns9,285 bytes
../../../../core/journal.jsruns13,631 bytes
../../../../core/loop-store.jsruns3,495 bytes
../../../../core/orchestrate.jsruns46,828 bytes
../../../../core/orientation.jsruns5,131 bytes
../../../../core/paths.jsruns8,834 bytes
../../../../core/project.jsruns6,797 bytes
../../../../core/prompts/index.jsruns38,033 bytes
../../../../core/provider.jsruns17,765 bytes
../../../../core/providers/antigravity.jsruns5,765 bytes
../../../../core/providers/codex.jsruns42,942 bytes
../../../../core/providers/grok.jsruns3,896 bytes
../../../../core/providers/openai-compatible.jsruns5,546 bytes
../../../../core/registry.jsruns9,963 bytes
../../../../core/result-cache.jsruns4,515 bytes
../../../../core/sessions.jsruns21,499 bytes
../../../../core/settings.jsruns2,778 bytes
../../../../core/signals.jsruns740 bytes
../../../../core/types.jsruns5,820 bytes
SKILL.mdcore16,377 bytes
SKILL.mdcore16,377 bytes
SKILL.mdcore16,377 bytes
../../../../scripts/commands/reload-mcp.shruns16,657 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes
../../../../server/dashboard/index.jsruns13,545 bytes
../../../../server/dashboard/runs.jsruns30,330 bytes
../../../../server/gemini/index.jsruns50,108 bytes
../../../../server/grok/index.jsruns64,158 bytes
../../../../server/mcp/index.jsruns131,046 bytes
../../../../server/mcp/index.jsruns131,046 bytes
../../../../server/mcp/index.jsruns131,046 bytes
../../../../server/mcp/index.jsruns131,046 bytes
../../../../server/mcp/index.jsruns131,046 bytes
../../../../server/openrouter/config.jsruns38,878 bytes
../../../../server/openrouter/index.jsruns27,271 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Deliberation

Host-neutral guidance for any AI coding agent connected to the deliberation MCP
server (Claude Code, Cursor, Codex, Kiro, Antigravity, Windsurf, Zed, and
others): what the tools do and when to delegate. The per-host skills and Powers
are generated from this file by `scripts/sync-hosts.js`.

## What deliberation is

A single MCP server that exposes GPT (via the Codex CLI), Gemini 3 (via the
Antigravity CLI), Grok (via the xAI API), local models (Ollama, LM Studio), and
OpenRouter models (400+, advisory) as expert subagents. You stay the primary
agent. When a task benefits from a second opinion or cross-model review, call

View the complete source on GitHub

plan-reviewer

When your agent uses it

Validate that a work plan is executable before work starts.

What it loads into context

Frontmatter

21 tokens

Always available to the agent

SKILL.md

814 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Modes
Default mode
Strict mode
Response Format
Modes of Operation
When to Invoke Plan Reviewer
When NOT to Invoke Plan Reviewer

Scripts executed separately

../../../../scripts/sync-hosts.js

Files

SKILL.mdcore3,730 bytes
SKILL.mdcore3,730 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Plan Reviewer

You are a work plan reviewer. You verify that a plan can actually be executed before anyone starts building.

## Context

You review a plan passed inline in the request. Each review is standalone. Your access varies by where you run: when you have filesystem or repo access, you may open referenced files to verify them; when you do not, judge whether references are named precisely enough to be found (exact path, function, doc section) rather than whether they exist on disk. Work from the context supplied and never assume details you have not actually seen.

## Modes

**Default - Blocker-only (approval bias):** You answer ONE question: "Can a capable developer execute this plan without getting stuck?" Approve when the plan is about 80% clear; a developer can resolve minor gaps. When in doubt, APPROVE.

**Strict:** Use this only when the request signals it - it contains "Review mode: strict", or the words strict / exhaustive / ruthless, or the plan is high-risk or architectural. In Strict mode you apply the full four-criteria rigor below and may list more issues.

View the complete source on GitHub

reload-mcp

When your agent uses it

Gracefully cycle dashboard and audit MCP processes after deliberation is updated.

What it loads into context

Frontmatter

24 tokens

Always available to the agent

SKILL.md

169 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Instructions

Scripts executed separately

../../../../scripts/commands/reload-mcp.sh
../../../../scripts/sync-hosts.js

Files

SKILL.mdcore844 bytes
SKILL.mdcore844 bytes
SKILL.mdcore844 bytes
../../../../scripts/commands/reload-mcp.shruns16,657 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Reload Deliberation MCP

Use this skill when Deliberation has been updated to cycle running background services and verify MCP worker states.

## Instructions

Run the reload script to audit running processes and gracefully restart the dashboard daemon on the latest code:

```bash
bash scripts/commands/reload-mcp.sh
```

- Preserves active host stdio pipes to prevent broken pipe errors.

View the complete source on GitHub

researcher

When your agent uses it

Research external libraries, APIs, and best practices, with evidence.

What it loads into context

Frontmatter

22 tokens

Always available to the agent

SKILL.md

736 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Capability Gate (read first)
Request Classification
Method
Response Format
Modes of Operation
When to Invoke Researcher
When NOT to Invoke Researcher

Scripts executed separately

../../../../scripts/sync-hosts.js

Files

SKILL.mdcore3,369 bytes
SKILL.mdcore3,369 bytes
SKILL.mdcore3,369 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Researcher

You are a research specialist for external libraries, frameworks, APIs, and open-source code. Your job: answer questions about third-party code with evidence, and stay honest about what you could and could not verify.

## Context

You operate as an on-demand specialist. Each consultation is standalone. Your available tools vary by where you run: some environments give you web search, documentation, repository, or shell access; others give you none. Adapt to what you actually have (capability gate below). Do not assume filesystem or repo access beyond what is provided.

## Capability Gate (read first)

- If you HAVE retrieval tools (web, docs, gh/git, code search): use them, then cite real, observed sources - URLs you fetched, GitHub permalinks with the commit SHA you saw, exact version numbers.
- If you do NOT have retrieval tools: answer from your own knowledge, but mark every non-trivial claim `[unverified]`, and NEVER fabricate links, commit SHAs, issue or PR numbers, version numbers, or API signatures. Instead, give the exact search or command the user could run to confirm (for example "search the official docs for X" or a `gh search code` query).
- Never present remembered detail as if it were freshly verified.

View the complete source on GitHub

scope-analyst

When your agent uses it

Catch ambiguities and hidden requirements before planning.

What it loads into context

Frontmatter

20 tokens

Always available to the agent

SKILL.md

1,163 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Phase 1: Intent Classification
Phase 2: Analysis
Anti-Patterns to Flag
Response Format
Modes of Operation
When to Invoke Scope Analyst
When NOT to Invoke Scope Analyst

Scripts executed separately

../../../../scripts/sync-hosts.js

Files

SKILL.mdcore5,442 bytes
SKILL.mdcore5,442 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Scope Analyst

You are a pre-planning consultant. Your job is to analyze requests BEFORE planning begins, catching ambiguities, hidden requirements, and pitfalls that would derail work later.

## Context

You operate at the earliest stage of the development workflow. Before anyone writes a plan or touches code, you make sure the request is fully understood. You prevent wasted effort by surfacing problems upfront. Your access varies by where you run: use filesystem or repo access when you have it, and when you do not, reason only from the context supplied. Never assume details you have not actually seen.

## Phase 1: Intent Classification

Classify intent FIRST, before any analysis. Every request maps to one type:

| Type | Focus | Key questions |

View the complete source on GitHub

security-analyst

When your agent uses it

Threat-model and find vulnerabilities, with practical remediation.

What it loads into context

Frontmatter

21 tokens

Always available to the agent

SKILL.md

934 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Analysis Framework
Response Format
Remediation Safety
Modes of Operation
Security Review Checklist
When to Invoke Security Analyst
When NOT to Invoke Security Analyst

Scripts executed separately

../../../../scripts/sync-hosts.js

Files

SKILL.mdcore4,328 bytes
SKILL.mdcore4,328 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Security Analyst

You are a security engineer specializing in application security, threat modeling, and vulnerability assessment.

## Context

You analyze code and systems with an attacker's mindset. Your job is to find vulnerabilities before attackers do, and to provide practical remediation - not theoretical concerns.

## Analysis Framework

### Threat Modeling

For any system or feature, identify:

View the complete source on GitHub

writing-editor

When your agent uses it

Edit prose so it reads like its author wrote it, keeping every fact and the author's voice.

What it loads into context

Frontmatter

28 tokens

Always available to the agent

SKILL.md

1,672 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Context
Method
What to flag
Keep (non-negotiable)
Honesty rule (important)
Response Format
Modes of Operation
When to Invoke
When NOT to Invoke

Scripts executed separately

../../../../scripts/sync-hosts.js

Files

SKILL.mdcore7,566 bytes
SKILL.mdcore7,566 bytes
../../../../scripts/sync-hosts.jsruns5,764 bytes

Read SKILL.md

<!-- GENERATED by scripts/sync-hosts.js - edit the source under prompts/, docs/tool-guide.md, or examples/, then regenerate. -->

# Writing Editor

You are a line editor for prose - posts, articles, newsletters, promo copy, READMEs, release notes. Your job is to make a draft read like its author wrote it: flag the machine-sounding patterns, keep every fact and the author's meaning, and return concrete line edits plus the rewritten text.

## Context

You are an on-demand advisor. Each consultation is standalone. You see the draft and whatever the author says about audience, purpose, and voice; when you have file access, read the draft and any voice file from disk, otherwise work from the text in the prompt. The draft may be one the author wrote or one an assistant just wrote for them; edit both the same way. You are an editor, not a detector: do not score the text for "AI likelihood", do not target any detection tool, and do not describe an edit as a way to pass one. Prose that sounds like a specific person is the goal; fooling a classifier is not.

## Method

1. Read the whole draft once for purpose, audience, and voice (register, sentence habits, vocabulary, language).
2. Establish the voice from evidence: a voice file, a writing sample, a description from the author, or the parts of the draft that clearly read as theirs. A voice file's style rules beat this prompt's style defaults; the Keep and Honesty sections below still apply. Use its signature moves only where their stated condition holds and within any stated limit, and treat its samples as rhythm, never as phrases to copy. If none of these exist, limit the pass to mechanical edits (filler, hedging, repetition) and say that voice fidelity could not be checked.
3. Flag the machine-sounding patterns below, each with the exact line or phrase it occurs in.

View the complete source on GitHub

Skills included

architect
code-reviewer
debugger
deliberation
plan-reviewer
reload-mcp
researcher
scope-analyst
security-analyst
writing-editor
architect
code-reviewer
debugger
deliberation
plan-reviewer
reload-mcp
researcher
scope-analyst
security-analyst
writing-editor

Repository facts

License: MIT
Created: May 13, 2026
Last pushed: Nov 4, 2026
Forks: 12
Open issues: 1
Project activity: Active

Before you install

Network or broad shell

Automated signals, not a guarantee

Checked Nov 6, 2026

Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.

Show 52 notes

.gemini/skills/architect/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
.gemini/skills/code-reviewer/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
.gemini/skills/debugger/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
.gemini/skills/deliberation/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
.gemini/skills/deliberation/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
core/prompts/index.js:7 Analysis incomplete: physical or joined line exceeds character limit
core/prompts/index.js:6 Analysis incomplete: physical or joined line exceeds character limit
core/prompts/index.js:9 Analysis incomplete: physical or joined line exceeds character limit
core/prompts/index.js:8 Analysis incomplete: physical or joined line exceeds character limit
core/prompts/index.js:11 [Credential-related source omitted]
core/prompts/index.js:10 Analysis incomplete: physical or joined line exceeds character limit
core/prompts/index.js:14 Analysis incomplete: physical or joined line exceeds character limit
core/prompts/index.js:12 Analysis incomplete: physical or joined line exceeds character limit
core/prompts/index.js:13 Analysis incomplete: physical or joined line exceeds character limit
server/mcp/index.js:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
server/mcp/index.js:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
server/mcp/index.js:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
server/mcp/index.js:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
.gemini/skills/plan-reviewer/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
.gemini/skills/reload-mcp/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed

No broad tool permissions
13 dangerous shell commandsin 4 files: scripts/commands/reload-mcp.sh:267 server/gemini/index.js:545 deliberation/SKILL.md:230 and 1 more
117 network referencesin 13 files: server/openrouter/index.js:69 server/gemini/index.js:624 server/mcp/index.js:1996 and 10 more
No hidden or obfuscated text
18 instruction override signalsin 2 files: server/gemini/index.js:185 core/providers/antigravity.js:49
102 executable scriptsin 35 files: scripts/sync-hosts.js:1 server/mcp/index.js:1 scripts/commands/reload-mcp.sh:1 and 32 more
64 references to secretsin 8 files: core/providers/codex.js:37 server/grok/index.js:26 core/providers/grok.js:33 and 5 more
No requests to vote in Hot

Show all 314 signals by file

`.claude-plugin/marketplace.json`

Network accessline 3

"owner": {     "name": "Anton Babenko",     "url": "https://github.com/antonbabenko"

Network accessline 4

"name": "Anton Babenko",     "url": "https://github.com/antonbabenko"   },

Network accessline 5

"url": "https://github.com/antonbabenko"

Network accessline 13

"author": {         "name": "Anton Babenko",         "url": "https://github.com/antonbabenko"

Network accessline 14

"name": "Anton Babenko",         "url": "https://github.com/antonbabenko"       },

Network accessline 15

"url": "https://github.com/antonbabenko"

Network accessline 16

},       "homepage": "https://github.com/antonbabenko/deliberation",       "repository": "https://github.com/antonbabenko/deliberation",

Network accessline 17

"homepage": "https://github.com/antonbabenko/deliberation",

Network accessline 18

"repository": "https://github.com/antonbabenko/deliberation",

`.claude-plugin/plugin.json`

Network accessline 5

"author": {     "name": "Anton Babenko",     "url": "https://github.com/antonbabenko"

Network accessline 6

"name": "Anton Babenko",     "url": "https://github.com/antonbabenko"   },

Network accessline 7

"url": "https://github.com/antonbabenko"

Network accessline 8

},   "homepage": "https://github.com/antonbabenko/deliberation",   "repository": "https://github.com/antonbabenko/deliberation",

Network accessline 9

"homepage": "https://github.com/antonbabenko/deliberation",

Network accessline 10

"repository": "https://github.com/antonbabenko/deliberation",

`.gemini/skills/deliberation/SKILL.md`

Network accessline 167

run to `<XDG cache>/deliberation/runs/` (override with `DELIBERATION_RUNS`), then run   `deliberation-mcp dashboard` (or `node server/mcp/index.js dashboard` from a checkout) and   open the printed `http://127.0.0.1:<port>/?t=<token>` URL.

Network accessline 168

`deliberation-mcp dashboard` (or `node server/mcp/index.js dashboard` from a checkout) and   open the printed `http://127.0.0.1:<port>/?t=<token>` URL. It is a read-only browser view   of live and past runs as state graphs, with config, pro

Network accessline 169

open the printed `http://127.0.0.1:<port>/?t=<token>` URL. It is a read-only browser view

Dangerous shell operationline 230

(`rm -rf ~/.npm/_npx`) or pin/refresh the version in your host's MCP config.

`core/analyze.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

Network accessline 56

const COMPARE_BASE = "https://openrouter.ai/compare";

Network accessline 56

const COMPARE_BASE = "https://openrouter.ai/compare";

`core/config-analysis.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/config-history.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/consensus-loop.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/debug-log.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/display.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/fanout-groups.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/host-budget.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/journal.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/loop-store.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/orchestrate.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/orientation.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/paths.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/project.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/prompts/index.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/provider.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

Network accessline 127

* a slow-but-alive provider was billed for a second full wait before failing anyway.  * @param {string} label  provider name for the message ("Grok", "OpenRouter")  * @param {any} err  the rejection from fetch()

Network accessline 127

* a slow-but-alive provider was billed for a second full wait before failing anyway.  * @param {string} label  provider name for the message ("Grok", "OpenRouter")  * @param {any} err  the rejection from fetch()

Network accessline 128

* @param {string} label  provider name for the message ("Grok", "OpenRouter")  * @param {any} err  the rejection from fetch()  * @param {number} timeoutMs  the bridge's own ceiling, for the abort message

Network accessline 128

* @param {string} label  provider name for the message ("Grok", "OpenRouter")  * @param {any} err  the rejection from fetch()  * @param {number} timeoutMs  the bridge's own ceiling, for the abort message

Network accessline 129

* @param {any} err  the rejection from fetch()

Network accessline 129

* @param {any} err  the rejection from fetch()

`core/providers/antigravity.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

Instruction override signalline 49

// Two-lock gate: write only when constructed write-capable AND this call explicitly asks.       // Anything else stays advisory (read-only). buildAgyArgs maps sandbox:"workspace-write" ->       // --dangerously-skip-permissions, else --san

Instruction override signalline 49

// Two-lock gate: write only when constructed write-capable AND this call explicitly asks.       // Anything else stays advisory (read-only). buildAgyArgs maps sandbox:"workspace-write" ->       // --dangerously-skip-permissions, else --san

Instruction override signalline 50

// Anything else stays advisory (read-only). buildAgyArgs maps sandbox:"workspace-write" ->       // --dangerously-skip-permissions, else --sandbox; runGemini's readOnly is the explicit       // caller decision (never argv-inferred) and eng

Instruction override signalline 50

// Anything else stays advisory (read-only). buildAgyArgs maps sandbox:"workspace-write" ->       // --dangerously-skip-permissions, else --sandbox; runGemini's readOnly is the explicit       // caller decision (never argv-inferred) and eng

Instruction override signalline 51

// --dangerously-skip-permissions, else --sandbox; runGemini's readOnly is the explicit

Instruction override signalline 51

// --dangerously-skip-permissions, else --sandbox; runGemini's readOnly is the explicit

`core/providers/codex.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```
Access to secretsline 37
```
[Credential-related source omitted]
```
Access to secretsline 37
```
[Credential-related source omitted]
```
Access to secretsline 186
```
[Credential-related source omitted]
```
Access to secretsline 186
```
[Credential-related source omitted]
```
Access to secretsline 187
```
[Credential-related source omitted]
```
Access to secretsline 187
```
[Credential-related source omitted]
```
Access to secretsline 188
```
[Credential-related source omitted]
```
Access to secretsline 188
```
[Credential-related source omitted]
```
Access to secretsline 189
```
[Credential-related source omitted]
```
Access to secretsline 189
```
[Credential-related source omitted]
```
Access to secretsline 199
```
[Credential-related source omitted]
```
Access to secretsline 199
```
[Credential-related source omitted]
```
Access to secretsline 200
```
[Credential-related source omitted]
```
Access to secretsline 200
```
[Credential-related source omitted]
```
Access to secretsline 205
```
[Credential-related source omitted]
```
Access to secretsline 205
```
[Credential-related source omitted]
```
Access to secretsline 206
```
[Credential-related source omitted]
```
Access to secretsline 206
```
[Credential-related source omitted]
```
Access to secretsline 207
```
[Credential-related source omitted]
```
Access to secretsline 207
```
[Credential-related source omitted]
```
Access to secretsline 216
```
[Credential-related source omitted]
```
Access to secretsline 216
```
[Credential-related source omitted]
```
Access to secretsline 237
```
[Credential-related source omitted]
```
Access to secretsline 237
```
[Credential-related source omitted]
```

`core/providers/grok.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

Network accessline 20

if (!bridge) throw new Error("makeGrokProvider requires opts.bridge (core is transport-agnostic; inject the grok bridge)");   const model = opts.model || process.env.GROK_DEFAULT_MODEL || "grok-4.6";   const apiBase = opts.apiBase || proces

Network accessline 20

if (!bridge) throw new Error("makeGrokProvider requires opts.bridge (core is transport-agnostic; inject the grok bridge)");   const model = opts.model || process.env.GROK_DEFAULT_MODEL || "grok-4.6";   const apiBase = opts.apiBase || proces

Network accessline 22

const apiBase = opts.apiBase || process.env.XAI_API_BASE || "https://api.x.ai/v1";

Network accessline 22

const apiBase = opts.apiBase || process.env.XAI_API_BASE || "https://api.x.ai/v1";

Access to secretsline 33
```
[Credential-related source omitted]
```
Access to secretsline 33
```
[Credential-related source omitted]
```
Access to secretsline 44
```
[Credential-related source omitted]
```
Access to secretsline 44
```
[Credential-related source omitted]
```

`core/providers/openai-compatible.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/registry.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/result-cache.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/sessions.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```
Access to secretsline 152
```
[Credential-related source omitted]
```
Access to secretsline 152
```
[Credential-related source omitted]
```
Access to secretsline 163
```
[Credential-related source omitted]
```
Access to secretsline 163
```
[Credential-related source omitted]
```

`core/settings.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/signals.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`core/types.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`plugins/deliberation/skills/deliberation/SKILL.md`

Network accessline 167

run to `<XDG cache>/deliberation/runs/` (override with `DELIBERATION_RUNS`), then run   `deliberation-mcp dashboard` (or `node server/mcp/index.js dashboard` from a checkout) and   open the printed `http://127.0.0.1:<port>/?t=<token>` URL.

Network accessline 168

`deliberation-mcp dashboard` (or `node server/mcp/index.js dashboard` from a checkout) and   open the printed `http://127.0.0.1:<port>/?t=<token>` URL. It is a read-only browser view   of live and past runs as state graphs, with config, pro

Network accessline 169

open the printed `http://127.0.0.1:<port>/?t=<token>` URL. It is a read-only browser view

Dangerous shell operationline 230

(`rm -rf ~/.npm/_npx`) or pin/refresh the version in your host's MCP config.

`scripts/commands/reload-mcp.sh`

Executable scriptline 1
```
#!/usr/bin/env bash
```
Executable scriptline 1
```
#!/usr/bin/env bash
```
Executable scriptline 1
```
#!/usr/bin/env bash
```
Executable scriptline 1
```
#!/usr/bin/env bash
```

Dangerous shell operationline 267

find "$HOME/.npm/_npx" -type d -path '*/@antonbabenko/deliberation-mcp' -prune -exec rm -rf {} + 2>/dev/null || true

Dangerous shell operationline 267

find "$HOME/.npm/_npx" -type d -path '*/@antonbabenko/deliberation-mcp' -prune -exec rm -rf {} + 2>/dev/null || true

Dangerous shell operationline 267

find "$HOME/.npm/_npx" -type d -path '*/@antonbabenko/deliberation-mcp' -prune -exec rm -rf {} + 2>/dev/null || true

Dangerous shell operationline 267

find "$HOME/.npm/_npx" -type d -path '*/@antonbabenko/deliberation-mcp' -prune -exec rm -rf {} + 2>/dev/null || true

Dangerous shell operationline 339

rm -f "$DASH_STATE_PATH" 2>/dev/null || true

Dangerous shell operationline 339

rm -f "$DASH_STATE_PATH" 2>/dev/null || true

Dangerous shell operationline 339

rm -f "$DASH_STATE_PATH" 2>/dev/null || true

Dangerous shell operationline 339

rm -f "$DASH_STATE_PATH" 2>/dev/null || true

`scripts/sync-hosts.js`

Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```

`server/dashboard/index.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

Network accessline 70

const urlLine = (port, token) => `Deliberation dashboard: http://127.0.0.1:${port}/?t=${token}\n`;

Network accessline 70

const urlLine = (port, token) => `Deliberation dashboard: http://127.0.0.1:${port}/?t=${token}\n`;

Network accessline 298

const url = `http://127.0.0.1:${bound}/?t=${token}`;

Network accessline 298

const url = `http://127.0.0.1:${bound}/?t=${token}`;

`server/dashboard/runs.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

`server/gemini/index.js`

Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```

Instruction override signalline 185

const readOnly = req.sandbox !== "workspace-write";   if (req.sandbox === "workspace-write") args.push("--dangerously-skip-permissions");

Instruction override signalline 185

const readOnly = req.sandbox !== "workspace-write";   if (req.sandbox === "workspace-write") args.push("--dangerously-skip-permissions");

Instruction override signalline 185

const readOnly = req.sandbox !== "workspace-write";   if (req.sandbox === "workspace-write") args.push("--dangerously-skip-permissions");

Instruction override signalline 186

if (req.sandbox === "workspace-write") args.push("--dangerously-skip-permissions");

Instruction override signalline 186

if (req.sandbox === "workspace-write") args.push("--dangerously-skip-permissions");

Instruction override signalline 186

if (req.sandbox === "workspace-write") args.push("--dangerously-skip-permissions");

Access to secretsline 221
```
[Credential-related source omitted]
```
Access to secretsline 221
```
[Credential-related source omitted]
```
Access to secretsline 221
```
[Credential-related source omitted]
```

Dangerous shell operationline 545

process.stderr.write("[deliberation] agy read-only run wrapped in sandbox-exec (workspace writes denied)\n");

Dangerous shell operationline 545

process.stderr.write("[deliberation] agy read-only run wrapped in sandbox-exec (workspace writes denied)\n");

Dangerous shell operationline 545

process.stderr.write("[deliberation] agy read-only run wrapped in sandbox-exec (workspace writes denied)\n");

Network accessline 624

settled = true;       if (err.code === "ENOENT") {         reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));

Network accessline 624

settled = true;       if (err.code === "ENOENT") {         reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));

Network accessline 624

settled = true;       if (err.code === "ENOENT") {         reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));

Network accessline 625

if (err.code === "ENOENT") {         reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));       } else {

Network accessline 625

if (err.code === "ENOENT") {         reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));       } else {

Network accessline 625

if (err.code === "ENOENT") {         reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));       } else {

Network accessline 626

reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));

Network accessline 626

reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));

Network accessline 626

reject(new Error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in."));

Instruction override signalline 889

const sandboxFlags = [];       if (args.sandbox === "workspace-write") sandboxFlags.push("--dangerously-skip-permissions");

Instruction override signalline 889

const sandboxFlags = [];       if (args.sandbox === "workspace-write") sandboxFlags.push("--dangerously-skip-permissions");

Instruction override signalline 889

const sandboxFlags = [];       if (args.sandbox === "workspace-write") sandboxFlags.push("--dangerously-skip-permissions");

Instruction override signalline 890

if (args.sandbox === "workspace-write") sandboxFlags.push("--dangerously-skip-permissions");

Instruction override signalline 890

if (args.sandbox === "workspace-write") sandboxFlags.push("--dangerously-skip-permissions");

Instruction override signalline 890

if (args.sandbox === "workspace-write") sandboxFlags.push("--dangerously-skip-permissions");

Network accessline 1044

execFileSync(AGY_CMD, [...AGY_PREFIX_ARGS, "--help"], { stdio: "ignore" });   } catch (e) {     console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");

Network accessline 1044

execFileSync(AGY_CMD, [...AGY_PREFIX_ARGS, "--help"], { stdio: "ignore" });   } catch (e) {     console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");

Network accessline 1044

execFileSync(AGY_CMD, [...AGY_PREFIX_ARGS, "--help"], { stdio: "ignore" });   } catch (e) {     console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");

Network accessline 1045

} catch (e) {     console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");     process.exit(1);

Network accessline 1045

} catch (e) {     console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");     process.exit(1);

Network accessline 1045

} catch (e) {     console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");     process.exit(1);

Network accessline 1046

console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");

Network accessline 1046

console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");

Network accessline 1046

console.error("Antigravity CLI (agy) not found. Install from https://antigravity.google and run `agy` once to sign in.");

`server/grok/index.js`

Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Network accessline 26
```
[Credential-related source omitted]
```
Network accessline 26
```
[Credential-related source omitted]
```
Network accessline 26
```
[Credential-related source omitted]
```
Access to secretsline 26
```
[Credential-related source omitted]
```
Access to secretsline 26
```
[Credential-related source omitted]
```
Access to secretsline 26
```
[Credential-related source omitted]
```

Network accessline 27

* > GROK_DEFAULT_MODEL (env) > grok-4.6.  * Endpoint: XAI_API_BASE (env) or https://api.x.ai/v1.  * File TTL: GROK_FILE_TTL_SECONDS (env) or 604800 (7 days).

Network accessline 27

* > GROK_DEFAULT_MODEL (env) > grok-4.6.  * Endpoint: XAI_API_BASE (env) or https://api.x.ai/v1.  * File TTL: GROK_FILE_TTL_SECONDS (env) or 604800 (7 days).

Network accessline 27

* > GROK_DEFAULT_MODEL (env) > grok-4.6.  * Endpoint: XAI_API_BASE (env) or https://api.x.ai/v1.  * File TTL: GROK_FILE_TTL_SECONDS (env) or 604800 (7 days).

Network accessline 28

* Endpoint: XAI_API_BASE (env) or https://api.x.ai/v1.

Network accessline 28

* Endpoint: XAI_API_BASE (env) or https://api.x.ai/v1.

Network accessline 28

* Endpoint: XAI_API_BASE (env) or https://api.x.ai/v1.

Network accessline 86

const DEFAULT_API_BASE = process.env.XAI_API_BASE || "https://api.x.ai/v1";

Network accessline 86

const DEFAULT_API_BASE = process.env.XAI_API_BASE || "https://api.x.ai/v1";

Network accessline 86

const DEFAULT_API_BASE = process.env.XAI_API_BASE || "https://api.x.ai/v1";

Access to secretsline 174
```
[Credential-related source omitted]
```
Access to secretsline 174
```
[Credential-related source omitted]
```
Access to secretsline 174
```
[Credential-related source omitted]
```
Access to secretsline 404
```
[Credential-related source omitted]
```
Access to secretsline 404
```
[Credential-related source omitted]
```
Access to secretsline 404
```
[Credential-related source omitted]
```
Access to secretsline 729
```
[Credential-related source omitted]
```
Access to secretsline 729
```
[Credential-related source omitted]
```
Access to secretsline 729
```
[Credential-related source omitted]
```
Access to secretsline 1257
```
[Credential-related source omitted]
```
Access to secretsline 1257
```
[Credential-related source omitted]
```
Access to secretsline 1257
```
[Credential-related source omitted]
```

Network accessline 1357

// Startup check: the bridge needs global fetch (Node 18+). The API key is NOT

Network accessline 1357

// Startup check: the bridge needs global fetch (Node 18+). The API key is NOT

Network accessline 1357

// Startup check: the bridge needs global fetch (Node 18+). The API key is NOT

Access to secretsline 1364
```
[Credential-related source omitted]
```
Access to secretsline 1364
```
[Credential-related source omitted]
```
Access to secretsline 1364
```
[Credential-related source omitted]
```
Access to secretsline 1365
```
[Credential-related source omitted]
```
Access to secretsline 1365
```
[Credential-related source omitted]
```
Access to secretsline 1365
```
[Credential-related source omitted]
```

`server/mcp/index.js`

Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
Executable file; content unavailable
```
Executable scriptline 1
```
Executable file; content unavailable
```
Executable scriptline 1
```
Executable file; content unavailable
```
Executable scriptline 1
```
Executable file; content unavailable
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
Executable file; content unavailable
```
Executable scriptline 1
```
Executable file; content unavailable
```
Executable scriptline 1
```
Executable file; content unavailable
```
Executable scriptline 1
```
Executable file; content unavailable
```
Access to secretsline 40
```
[Credential-related source omitted]
```
Access to secretsline 40
```
[Credential-related source omitted]
```
Access to secretsline 40
```
[Credential-related source omitted]
```

Network accessline 1996

makeOpenAICompatibleProvider({       name: "ollama",       apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",

Network accessline 1996

makeOpenAICompatibleProvider({       name: "ollama",       apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",

Network accessline 1996

makeOpenAICompatibleProvider({       name: "ollama",       apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",

Network accessline 1997

name: "ollama",       apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",       apiKeyEnv: (initialProviders.ollama && initialProviders.ollama.apiKeyEnv) || "",

Network accessline 1997

name: "ollama",       apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",       apiKeyEnv: (initialProviders.ollama && initialProviders.ollama.apiKeyEnv) || "",

Network accessline 1997

name: "ollama",       apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",       apiKeyEnv: (initialProviders.ollama && initialProviders.ollama.apiKeyEnv) || "",

Network accessline 1998

apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",

Network accessline 1998

apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",

Network accessline 1998

apiBase: (initialProviders.ollama && initialProviders.ollama.apiBase) || "http://localhost:11434/v1",

Network accessline 2004

makeOpenAICompatibleProvider({       name: "lmstudio",       apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",

Network accessline 2004

makeOpenAICompatibleProvider({       name: "lmstudio",       apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",

Network accessline 2004

makeOpenAICompatibleProvider({       name: "lmstudio",       apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",

Network accessline 2005

name: "lmstudio",       apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",       apiKeyEnv: (initialProviders.lmstudio && initialProviders.lmstudio.apiKeyEnv) || "",

Network accessline 2005

name: "lmstudio",       apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",       apiKeyEnv: (initialProviders.lmstudio && initialProviders.lmstudio.apiKeyEnv) || "",

Network accessline 2005

name: "lmstudio",       apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",       apiKeyEnv: (initialProviders.lmstudio && initialProviders.lmstudio.apiKeyEnv) || "",

Network accessline 2006

apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",

Network accessline 2006

apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",

Network accessline 2006

apiBase: (initialProviders.lmstudio && initialProviders.lmstudio.apiBase) || "http://localhost:1234/v1",

`server/openrouter/config.js`

Executable scriptline 1
```
Executable file
```
Executable scriptline 1
```
Executable file
```

Network accessline 14

const DEFAULT_API_BASE = "https://openrouter.ai/api/v1";

Network accessline 14

const DEFAULT_API_BASE = "https://openrouter.ai/api/v1";

Access to secretsline 15
```
[Credential-related source omitted]
```
Access to secretsline 15
```
[Credential-related source omitted]
```

`server/openrouter/index.js`

Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```
Executable scriptline 1
```
#!/usr/bin/env node
```

Network accessline 69

const f = fetchImpl || globalThis.fetch;   if (typeof f !== "function") { const e = new Error("global fetch unavailable; Node 18+ required"); e.code = "network"; throw e; }   const base = (apiBase || "https://openrouter.ai/api/v1").replace(

Network accessline 69

const f = fetchImpl || globalThis.fetch;   if (typeof f !== "function") { const e = new Error("global fetch unavailable; Node 18+ required"); e.code = "network"; throw e; }   const base = (apiBase || "https://openrouter.ai/api/v1").replace(

Network accessline 69

const f = fetchImpl || globalThis.fetch;   if (typeof f !== "function") { const e = new Error("global fetch unavailable; Node 18+ required"); e.code = "network"; throw e; }   const base = (apiBase || "https://openrouter.ai/api/v1").replace(

Network accessline 70

if (typeof f !== "function") { const e = new Error("global fetch unavailable; Node 18+ required"); e.code = "network"; throw e; }   const base = (apiBase || "https://openrouter.ai/api/v1").replace(/\/+$/, "");   const url = `${base}/chat/co

Network accessline 70

if (typeof f !== "function") { const e = new Error("global fetch unavailable; Node 18+ required"); e.code = "network"; throw e; }   const base = (apiBase || "https://openrouter.ai/api/v1").replace(/\/+$/, "");   const url = `${base}/chat/co

Network accessline 70

if (typeof f !== "function") { const e = new Error("global fetch unavailable; Node 18+ required"); e.code = "network"; throw e; }   const base = (apiBase || "https://openrouter.ai/api/v1").replace(/\/+$/, "");   const url = `${base}/chat/co

Network accessline 71

const base = (apiBase || "https://openrouter.ai/api/v1").replace(/\/+$/, "");

Network accessline 71

const base = (apiBase || "https://openrouter.ai/api/v1").replace(/\/+$/, "");

Network accessline 71

const base = (apiBase || "https://openrouter.ai/api/v1").replace(/\/+$/, "");

Network accessline 80

const headers = {     "Content-Type": "application/json",     "HTTP-Referer": "https://github.com/antonbabenko/deliberation",

Network accessline 80

const headers = {     "Content-Type": "application/json",     "HTTP-Referer": "https://github.com/antonbabenko/deliberation",

Network accessline 80

const headers = {     "Content-Type": "application/json",     "HTTP-Referer": "https://github.com/antonbabenko/deliberation",

Network accessline 81

"Content-Type": "application/json",     "HTTP-Referer": "https://github.com/antonbabenko/deliberation",     "X-Title": "deliberation",

Network accessline 81

"Content-Type": "application/json",     "HTTP-Referer": "https://github.com/antonbabenko/deliberation",     "X-Title": "deliberation",

Network accessline 81

"Content-Type": "application/json",     "HTTP-Referer": "https://github.com/antonbabenko/deliberation",     "X-Title": "deliberation",

Network accessline 82

"HTTP-Referer": "https://github.com/antonbabenko/deliberation",

Network accessline 82

"HTTP-Referer": "https://github.com/antonbabenko/deliberation",

Network accessline 82

"HTTP-Referer": "https://github.com/antonbabenko/deliberation",

Network accessline 193

*/ function resolveEffectiveAuth(delegate, or) {   const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";

Network accessline 193

*/ function resolveEffectiveAuth(delegate, or) {   const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";

Network accessline 193

*/ function resolveEffectiveAuth(delegate, or) {   const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";

Network accessline 194

function resolveEffectiveAuth(delegate, or) {   const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";   const providerKeyEnv = or.apiKeyEnv || "OPENROUTER_API_KEY";

Network accessline 194

function resolveEffectiveAuth(delegate, or) {   const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";   const providerKeyEnv = or.apiKeyEnv || "OPENROUTER_API_KEY";

Network accessline 194

function resolveEffectiveAuth(delegate, or) {   const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";   const providerKeyEnv = or.apiKeyEnv || "OPENROUTER_API_KEY";

Network accessline 195

const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";

Network accessline 195

const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";

Network accessline 195

const providerApiBase = or.apiBase || "https://openrouter.ai/api/v1";

Access to secretsline 196
```
[Credential-related source omitted]
```
Access to secretsline 196
```
[Credential-related source omitted]
```
Access to secretsline 196
```
[Credential-related source omitted]
```

How the signals fit together

Facts from a static read of the files at the pinned commit. They describe what the code does, not why.

Some files could not be read; the facts below still hold.

Declared and executed: 7 facts

Contacts openrouter.ai, which SKILL.md, the README and the plugin manifest do not mentioncore/analyze.js:56High confidence
Contacts api.x.ai, which SKILL.md, the README and the plugin manifest do not mentioncore/providers/grok.js:22High confidence
Contacts antigravity.google, which SKILL.md, the README and the plugin manifest do not mentionserver/gemini/index.js:626High confidence
Contacts api.x.ai, which SKILL.md, the README and the plugin manifest do not mentionserver/grok/index.js:86High confidence
Contacts openrouter.ai, which SKILL.md, the README and the plugin manifest do not mentionserver/openrouter/config.js:14High confidence
Contacts openrouter.ai, which SKILL.md, the README and the plugin manifest do not mentionserver/openrouter/index.js:71High confidence
Contacts github.com, which SKILL.md, the README and the plugin manifest do not mentionserver/openrouter/index.js:82High confidence

Signal confidence: 299 facts

Network access in prose of the SKILL.mddeliberation/SKILL.md:167Low confidence
Network access in prose of the SKILL.mddeliberation/SKILL.md:168Low confidence
Network access in prose of the SKILL.mddeliberation/SKILL.md:169Low confidence
Dangerous shell operation in prose of the SKILL.mddeliberation/SKILL.md:230Low confidence
Executable script in code of a file that runscore/analyze.js:1High confidence
Executable script in code of a file that runscore/analyze.js:1High confidence
Network access in code of a file that runscore/analyze.js:56High confidence
Network access in code of a file that runscore/analyze.js:56High confidence
Executable script in code of a file that runscore/config-analysis.js:1High confidence
Executable script in code of a file that runscore/config-analysis.js:1High confidence
Executable script in code of a file that runscore/config-history.js:1High confidence
Executable script in code of a file that runscore/config-history.js:1High confidence
Executable script in code of a file that runscore/consensus-loop.js:1High confidence
Executable script in code of a file that runscore/consensus-loop.js:1High confidence
Executable script in code of a file that runscore/debug-log.js:1High confidence
Executable script in code of a file that runscore/debug-log.js:1High confidence
Executable script in code of a file that runscore/display.js:1High confidence
Executable script in code of a file that runscore/display.js:1High confidence
Executable script in code of a file that runscore/fanout-groups.js:1High confidence
Executable script in code of a file that runscore/fanout-groups.js:1High confidence
Executable script in code of a file that runscore/host-budget.js:1High confidence
Executable script in code of a file that runscore/host-budget.js:1High confidence
Executable script in code of a file that runscore/journal.js:1High confidence
Executable script in code of a file that runscore/journal.js:1High confidence
Executable script in code of a file that runscore/loop-store.js:1High confidence
Executable script in code of a file that runscore/loop-store.js:1High confidence
Executable script in code of a file that runscore/orchestrate.js:1High confidence
Executable script in code of a file that runscore/orchestrate.js:1High confidence
Executable script in code of a file that runscore/orientation.js:1High confidence
Executable script in code of a file that runscore/orientation.js:1High confidence
Executable script in code of a file that runscore/paths.js:1High confidence
Executable script in code of a file that runscore/paths.js:1High confidence
Executable script in code of a file that runscore/project.js:1High confidence
Executable script in code of a file that runscore/project.js:1High confidence
Executable script in code of a file that runscore/prompts/index.js:1High confidence
Executable script in code of a file that runscore/prompts/index.js:1High confidence
Executable script in code of a file that runscore/provider.js:1High confidence
Executable script in code of a file that runscore/provider.js:1High confidence
Network access in a comment of a file that runscore/provider.js:127Low confidence
Network access in a comment of a file that runscore/provider.js:127Low confidence
Network access in a comment of a file that runscore/provider.js:128Low confidence
Network access in a comment of a file that runscore/provider.js:128Low confidence
Network access in a comment of a file that runscore/provider.js:129Low confidence
Network access in a comment of a file that runscore/provider.js:129Low confidence
Executable script in code of a file that runscore/providers/antigravity.js:1High confidence
Executable script in code of a file that runscore/providers/antigravity.js:1High confidence
Instruction override signal in a comment of a file that runscore/providers/antigravity.js:49Low confidence
Instruction override signal in a comment of a file that runscore/providers/antigravity.js:49Low confidence
Instruction override signal in a comment of a file that runscore/providers/antigravity.js:50Low confidence
Instruction override signal in a comment of a file that runscore/providers/antigravity.js:50Low confidence

and 249 more

Works with

Agent compatibility, from the files in this repository
Agent	Status	Evidence	Install to
Claude Code	Native	Manifest .claude-plugin/marketplace.json	`.claude/skills/<skill-name>/`
Codex	Native	Manifest AGENTS.md	`.agents/skills/<skill-name>/`
Cursor	Native	Path .cursor/rules/deliberation.mdc	`.cursor/skills/<skill-name>/`
Gemini CLI	Native	Path .gemini/skills/architect/SKILL.md	`.gemini/skills/<skill-name>/`
GitHub Copilot	Compatible	Standard SKILL.md .gemini/skills/architect/SKILL.md	`.github/skills/<skill-name>/`
OpenCode	Native	Path .opencode/agents/architect.md	`.opencode/skills/<skill-name>/`

Native The repository ships a manifest, folder or frontmatter field for this agent.
Compatible A standard SKILL.md that this agent can read.
Unknown No evidence either way.

Derived automatically from paths, manifests and frontmatter. Not tested on each agent.