DevOtts/build-it

## Key results - build-it v3.1.0 released: repo renamed on GitHub (redirects live), 35-file rename commit 1d78426, lints 9/9 green. https://github.com/DevOtts/build-it/releases/tag/v3.1.0

- build-it v3.1.0 released: repo renamed on GitHub (redirects live), 35-file rename commit 1d78426, lints 9/9 green. https://github.com/DevOtts/build-it/releases/tag/v3.1.0

**Pointers:** .build-it-reports/report-v3.1.1-assets.md · https://github.com/DevOtts/build-it/releases/tag/v3.1.1 ·

https://github.com/DevOtts/build-it/releases/tag/v3.1.1 ·

"owner": {     "name": "DevOtts",     "url": "https://github.com/DevOtts"

"name": "DevOtts",     "url": "https://github.com/DevOtts"   },

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

"author": {         "name": "DevOtts",         "url": "https://github.com/DevOtts"

"name": "DevOtts",         "url": "https://github.com/DevOtts"       },

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

},       "homepage": "https://github.com/DevOtts/prompt-it",       "license": "MIT",

"homepage": "https://github.com/DevOtts/prompt-it",

"author": {         "name": "DevOtts",         "url": "https://github.com/DevOtts"

"name": "DevOtts",         "url": "https://github.com/DevOtts"       }

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

[Credential-related source omitted]

All notable changes to build-it (formerly fable-it) are documented here. The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/);

format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/);

versions follow [Semantic Versioning](https://semver.org/).

[3.0.1]: https://github.com/DevOtts/fable-it/releases/tag/v3.0.1

[3.0.0]: https://github.com/DevOtts/fable-it/releases/tag/v3.0.0

[2.1.0]: https://github.com/DevOtts/fable-it/releases/tag/v2.1.0

[2.0.0]: https://github.com/DevOtts/fable-it/releases/tag/v2.0.0

[0.1.0]: https://github.com/DevOtts/fable-it/commits/cb1b17a

Copyright (c) 2026 DevOtts (https://github.com/DevOtts)

<p>   <a href="#installation"><img src="https://img.shields.io/badge/Quick%20Start-Install-blue?style=for-the-badge" alt="Quick Start"></a>

<a href="#installation"><img src="https://img.shields.io/badge/Quick%20Start-Install-blue?style=for-the-badge" alt="Quick Start"></a>

<a href="https://opensource.org/licenses/MIT"><img src="https://img.shields.io/badge/License-MIT-yellow.svg?style=for-the-badge" alt="License: MIT"></a>

<a href="https://claude.ai/code"><img src="https://img.shields.io/badge/Works%20with-Claude%20Code-orange?style=for-the-badge&logo=anthropic" alt="Works with Claude Code"></a>

<a href="https://claude.ai/code"><img src="https://img.shields.io/badge/tuned%20for-Sonnet%205-blueviolet?style=for-the-badge" alt="Tuned for Sonnet 5"></a>

<a href="https://claude.ai/code"><img src="https://img.shields.io/badge/tuned%20for-Opus%204.8-blueviolet?style=for-the-badge" alt="Tuned for Opus 4.8"></a>

<a href="https://github.com/DevOtts"><img src="https://img.shields.io/badge/author-DevOtts-181717?style=for-the-badge&logo=github" alt="Author: DevOtts"></a>

> [!NOTE] > **The lifecycle family** — [`plan-it`](https://github.com/DevOtts/plan-it) plans → **build-it** builds → [`review-it`](https://github.com/DevOtts/review-it) verifies. review-it is the QA front door: it runs the plan-phase Test C

> **The lifecycle family** — [`plan-it`](https://github.com/DevOtts/plan-it) plans → **build-it** builds → [`review-it`](https://github.com/DevOtts/review-it) verifies. review-it is the QA front door: it runs the plan-phase Test Contract ag

git clone https://github.com/DevOtts/build-it /tmp/build-it

build-it is a `SKILL.md` at its core, so it runs anywhere the open [Agent Skills](https://github.com/vercel-labs/skills) standard does. **Claude Code** gets the full plugin (conductor **+** the four bundled skills); every other tool gets th

Thank you for contribution: - [andre2654](https://github.com/andre2654)

- [andre2654](https://github.com/andre2654)

- [mBidarra](https://github.com/mBidarra)

- [her0 Studio](https://github.com/her0-studio/)

_Built by [DevOtts](https://github.com/DevOtts)._

license: MIT author: DevOtts author_url: https://github.com/DevOtts

author: DevOtts author_url: https://github.com/DevOtts homepage: https://github.com/DevOtts/build-it

author_url: https://github.com/DevOtts

homepage: https://github.com/DevOtts/build-it

repository: https://github.com/DevOtts/build-it

--- _Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

for criterion 2 (the sync handler code) is completed anyway. The run reaches the report phase. No entry for criterion 2 exists in `.taskstate/evidence.md` (the only attempted probe — the timed-out curl — is logged there as a failed reachabi

report phase. No entry for criterion 2 exists in `.taskstate/evidence.md` (the only attempted probe — the timed-out curl — is logged there as a failed reachability check, not a passing verification).

attempted probe — the timed-out curl — is logged there as a failed reachability

2. The verifiability precheck in the cycle phase routes criterion 2 straight to    IMPLEMENTED-NOT-VERIFIED with the reason ("staging tenant unreachable; curl    timed out; built to spec, not verified") — no QA executor is spawned against a

IMPLEMENTED-NOT-VERIFIED with the reason ("staging tenant unreachable; curl    timed out; built to spec, not verified") — no QA executor is spawned against a    mock, no mock result is treated as verification.

Two fable-it QA runs execute in parallel on the same machine. - Run A uses the defaults: no `CDP_URL` set, app under test on the default app port. - Run B was started with `CDP_URL=http://localhost:9223` and its app on port 3100

- Run A uses the defaults: no `CDP_URL` set, app under test on the default app port. - Run B was started with `CDP_URL=http://localhost:9223` and its app on port 3100   (recorded in Run B's `.taskstate/grounding.md` / environment).

- Run B was started with `CDP_URL=http://localhost:9223` and its app on port 3100

1. Each run resolves its CDP endpoint through the shared core's resolution order    (env `CDP_URL` → grounding/test-plan value → default) instead of a hardcoded    URL: Run A resolves the default; Run B resolves `http://localhost:9223`.

(env `CDP_URL` → grounding/test-plan value → default) instead of a hardcoded    URL: Run A resolves the default; Run B resolves `http://localhost:9223`. 2. App URLs are likewise resolved from env/grounding/test plan, not hardcoded:

URL: Run A resolves the default; Run B resolves `http://localhost:9223`.

**Released:** merged to main (b300d2a), tagged v3.0.1, published https://github.com/DevOtts/fable-it/releases/tag/v3.0.1 (2026-07-09). Prod-verified: fresh clone of the tag — 9 lints + 7 hook tests green, plugin+marketplace validate pass, n

replacing it. The existing `SessionStart` only has a `compact` matcher; add a    separate startup entry. 2. Make `/full-qa` read `${CDP_URL:-http://localhost:9222}` and `$PORT` from

separate startup entry. 2. Make `/full-qa` read `${CDP_URL:-http://localhost:9222}` and `$PORT` from    `.env.worktree` in preflight, keeping `:9222` only as the explicit model-3

2. Make `/full-qa` read `${CDP_URL:-http://localhost:9222}` and `$PORT` from

`PreToolUse` Bash hook (`rtk hook claude`). MERGE the new SessionStart/   SessionEnd entries; do not overwrite this block. - `/full-qa` hardcodes `http://localhost:9222` in three places (lines ~51, ~127,

SessionEnd entries; do not overwrite this block. - `/full-qa` hardcodes `http://localhost:9222` in three places (lines ~51, ~127,   ~367) and uses `localhost:PORT` placeholders. That is the actual collision.

- `/full-qa` hardcodes `http://localhost:9222` in three places (lines ~51, ~127,

| # | Source | URL | Class | What it gave us | |---|--------|-----|-------|-----------------| | 1 | Anthropic announcement: "Claude Fable 5 and Claude Mythos 5" | https://www.anthropic.com/news/claude-fable-5-mythos-5 | a / c | Long-horizon

|---|--------|-----|-------|-----------------| | 1 | Anthropic announcement: "Claude Fable 5 and Claude Mythos 5" | https://www.anthropic.com/news/claude-fable-5-mythos-5 | a / c | Long-horizon framing, Stripe case study, Slay-the-Spire mem

| 1 | Anthropic announcement: "Claude Fable 5 and Claude Mythos 5" | https://www.anthropic.com/news/claude-fable-5-mythos-5 | a / c | Long-horizon framing, Stripe case study, Slay-the-Spire memory result, pricing, system-card link |

| 2 | Platform docs: "Introducing Claude Fable 5 and Claude Mythos 5" | https://platform.claude.com/docs/en/about-claude/models/introducing-claude-fable-5-and-claude-mythos-5 | a | 1M context, 128k output, adaptive thinking, supported tools

| 3 | Platform docs: **"Prompting Claude Fable 5"** | https://platform.claude.com/docs/en/build-with-claude/prompt-engineering/prompting-claude-fable-5 | a | **The keystone source.** Fable's behavioral deltas vs Opus 4.8 + exact portable pr

| 4 | Claude Fable 5 & Mythos 5 System Card (PDF, ~319pp) | https://www-cdn.anthropic.com/2f9323abbcc4abe219577539efe19a623c9ca2bd/Claude%20Fable%205%20&%20Claude%20Mythos%205%20System%20Card.pdf | a | Could not fetch directly (>10MB). Cont

| 5 | Simon Willison — "Initial impressions of Claude Fable 5" | https://simonwillison.net/2026/Jun/9/claude-fable-5/ | d | Proactivity, scope-expansion behavior, model-size proxy, "relentlessly proactive" |

| 6 | Vellum — "Fable 5 & Mythos 5 benchmarks explained" | https://www.vellum.ai/blog/claude-fable-5-and-mythos-5-benchmarks-explained | b / d | SWE-Bench Pro, FrontierCode Diamond, vision, bio/cyber numbers |

| 7 | claude5.ai — "80.3% on SWE-Bench Pro" | https://claude5.ai/en/news/claude-fable-5-benchmarks-swe-bench-pro-80-percent | b / d | SWE-Bench Pro head-to-head vs Opus/GPT/Gemini |

| 8 | Zvi Mowshowitz — "Fable 5 and Mythos 5: The System Card" | https://thezvi.substack.com/p/claude-fable-5-and-mythos-5-the-system | d | Relayed system-card numbers on hallucinated references, eval awareness, code-review honesty |

| 9 | Claude Code changelog | https://code.claude.com/docs/en/changelog | a | ultracode effort tier, agent-teams fixes, compaction/1M-context fixes, subagent thinking inheritance |

| 10 | Anthropic — "Introducing Claude Sonnet 5" | https://www.anthropic.com/news/claude-sonnet-5 | a | Confirms Sonnet 5 (2026-06-30) as a co-target model for fable-it |

| 11 | InfoQ — Fable 5 release/suspension news | https://www.infoq.com/news/2026/06/claude-5-release/ | d | Timeline confirmation (release, export-control suspension, restoration) |

1. **Fable-specific METR time-horizon number** (the "task length at 50% success" minutes/hours figure). Only METR's general methodology surfaced. If a Fable time-horizon figure exists it's likely inside the 319-page system card, which I cou

ran `mv ~/Workspace/Devotts/fable-it ~/Workspace/Devotts/build-it`; if the   folder is still `fable-it`, stop and say so). - Remote: https://github.com/DevOtts/build-it, branch `main`, released tag

folder is still `fable-it`, stop and say so). - Remote: https://github.com/DevOtts/build-it, branch `main`, released tag   `v3.1.0` (rename commit `1d78426`).

- Remote: https://github.com/DevOtts/build-it, branch `main`, released tag

"author": {     "name": "DevOtts",     "url": "https://github.com/DevOtts"

"name": "DevOtts",     "url": "https://github.com/DevOtts"   },

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

},   "homepage": "https://github.com/DevOtts/build-it",   "repository": "https://github.com/DevOtts/build-it",

"homepage": "https://github.com/DevOtts/build-it",

"repository": "https://github.com/DevOtts/build-it",

--- _Part of the build-it plugin · authored by [DevOtts](https://github.com/DevOtts)._

_Part of the build-it plugin · authored by [DevOtts](https://github.com/DevOtts)._

#!/usr/bin/env python3

[Credential-related source omitted]

#!/usr/bin/env bash

trap 'rm -rf "$TMP"' EXIT

#!/usr/bin/env python3

[Credential-related source omitted]

#!/usr/bin/env node

#!/usr/bin/env node

--- _Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

Executable file

// run it with a timeout in the repo root, and look up the (task, git sha, command) cache. // Nothing here runs on `view` or `validate`; only the explicit `recheck` command calls runCheck. // Pure ESM, zero dependencies. Authored by DevOtts

// Pure ESM, zero dependencies. Authored by DevOtts (https://github.com/DevOtts).

// status-lib.mjs - schema 1 status.json: constants, validator, view, atomic write, lock. // Pure ESM, zero dependencies. Authored by DevOtts (https://github.com/DevOtts).

Executable file

// Pure ESM, zero dependencies. Authored by DevOtts (https://github.com/DevOtts).

#!/usr/bin/env node

Executable file

#!/usr/bin/env node

#!/usr/bin/env node

// status-write.mjs - the single writer for status.json (schema 1). Zero dependencies. // Exit codes: 0 ok | 1 invalid or refused by rule | 2 usage error / file exists | 3 lock or concurrency failure. // Authored by DevOtts (https://github.

// status-write.mjs - the single writer for status.json (schema 1). Zero dependencies. // Exit codes: 0 ok | 1 invalid or refused by rule | 2 usage error / file exists | 3 lock or concurrency failure. // Authored by DevOtts (https://github.

// Authored by DevOtts (https://github.com/DevOtts).

// Authored by DevOtts (https://github.com/DevOtts).

{   "$schema": "https://json-schema.org/draft/2020-12/schema",

"$schema": "https://json-schema.org/draft/2020-12/schema",

"$id": "https://github.com/DevOtts/build-it/status.schema.json",

description: Single-command autonomous delivery orchestrator. Invoke it with a goal and a numbered Definition of Done (DoD) and it runs the whole job to completion — typically unattended, overnight — by conducting the bundled /launch, /iter

author: DevOtts author_url: https://github.com/DevOtts ---

author_url: https://github.com/DevOtts

--- _Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

can be solved by an API call to the same service. author: DevOtts author_url: https://github.com/DevOtts

author: DevOtts author_url: https://github.com/DevOtts ---

author_url: https://github.com/DevOtts

--- _Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

description: Generic autonomous QA suite for any project — reads a test plan (markdown file or inline spec), runs all tests using Chrome CDP + iterate cycles, fixes bugs found, and produces a final QA report. Works with any web stack. Trigg

author: DevOtts author_url: https://github.com/DevOtts ---

author_url: https://github.com/DevOtts

curl -s -o /dev/null -w "%{http_code}" <SERVICE_URL>/health

curl -s -o /dev/null -w "%{http_code}" <SERVICE_URL>

curl -s "$CDP_URL/json/version" | python3 -c "import sys,json; d=json.load(sys.stdin); print('Chrome:', d.get('Browser','?'))"

curl -s "$CDP_URL/json/version" | python3 -c "import sys,json; d=json.load(sys.stdin); print('Chrome:', d.get('Browser','?'))"

PREFLIGHT   ✓ Service A  http://localhost:XXXX  200

✓ Service A  http://localhost:XXXX  200

✓ Service B  http://localhost:YYYY  200

✗ Service C  http://localhost:ZZZZ  ECONNREFUSED  ← fixing...

### API-only tests (curl + DB)

docker exec -i supabase_db_supabase psql -U postgres < migration.sql # Auth token curl -s -X POST "http://127.0.0.1:54321/auth/v1/token?grant_type=password" \

[Credential-related source omitted]

curl -s -X POST "http://127.0.0.1:54321/auth/v1/token?grant_type=password" \

# Health curl -s http://localhost:PORT/health

curl -s http://localhost:PORT/health

# Health curl -s http://127.0.0.1:PORT/health

curl -s http://127.0.0.1:PORT/health

--- _Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

description: Autonomous multi-cycle problem solver for complex tasks that require diagnosis → fix → test → verify loops. Use when the user says things like "make this work", "test this end-to-end", "fix and verify", "iterate until working",

author: DevOtts author_url: https://github.com/DevOtts ---

author_url: https://github.com/DevOtts

--- _Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

description: Mission control for autonomous projects — analyzes tasks, recommends approaches (sub-agents/teams), sets up environment (features, evals, hooks, init.sh), suggests and configures tooling (MCPs), and launches the work. author: D

author: DevOtts author_url: https://github.com/DevOtts ---

author_url: https://github.com/DevOtts

[Credential-related source omitted]

│             2. Add to .env: │                LANGSMITH_TRACING=true │                LANGSMITH_ENDPOINT=https://api.smith.langchain.com

[Credential-related source omitted]

│                LANGSMITH_ENDPOINT=https://api.smith.langchain.com

[Credential-related source omitted]

[Credential-related source omitted]

echo "Waiting for server..."   for i in $(seq 1 30); do     if curl -s http://localhost:${PORT:-3000}/api/health > /dev/null 2>&1; then

for i in $(seq 1 30); do     if curl -s http://localhost:${PORT:-3000}/api/health > /dev/null 2>&1; then       echo "Server ready on :${PORT:-3000}"

if curl -s http://localhost:${PORT:-3000}/api/health > /dev/null 2>&1; then

claude -p "$(cat .claude/prompts/launch-[project].md)" \   --dangerously-skip-permissions

--dangerously-skip-permissions

--- _Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

1. `CDP_URL` environment variable 2. a CDP endpoint recorded in `.taskstate/grounding.md` or the test plan 3. the default: `http://localhost:9222` — this line is the default declaration, the only place the port is written

1. `CDP_URL` environment variable 2. a CDP endpoint recorded in `.taskstate/grounding.md` or the test plan 3. the default: `http://localhost:9222` — this line is the default declaration, the only place the port is written

1. `CDP_URL` environment variable 2. a CDP endpoint recorded in `.taskstate/grounding.md` or the test plan 3. the default: `http://localhost:9222` — this line is the default declaration, the only place the port is written

3. the default: `http://localhost:9222` — this line is the default declaration, the only place the port is written

3. the default: `http://localhost:9222` — this line is the default declaration, the only place the port is written

3. the default: `http://localhost:9222` — this line is the default declaration, the only place the port is written

curl -s "${CDP_URL:-http://localhost:9222}/json/version"   # ${...:-} carries the default

curl -s "${CDP_URL:-http://localhost:9222}/json/version"   # ${...:-} carries the default

curl -s "${CDP_URL:-http://localhost:9222}/json/version"   # ${...:-} carries the default

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

[Credential-related source omitted]

if page is None:         page = await ctx.new_page()         await page.goto("https://TARGET_URL", wait_until="domcontentloaded", timeout=30000)

if page is None:         page = await ctx.new_page()         await page.goto("https://TARGET_URL", wait_until="domcontentloaded", timeout=30000)

if page is None:         page = await ctx.new_page()         await page.goto("https://TARGET_URL", wait_until="domcontentloaded", timeout=30000)

await page.goto("https://TARGET_URL", wait_until="domcontentloaded", timeout=30000)

await page.goto("https://TARGET_URL", wait_until="domcontentloaded", timeout=30000)

await page.goto("https://TARGET_URL", wait_until="domcontentloaded", timeout=30000)

--- _Part of the build-it plugin · authored by [DevOtts](https://github.com/DevOtts)._

--- _Part of the build-it plugin · authored by [DevOtts](https://github.com/DevOtts)._

--- _Part of the build-it plugin · authored by [DevOtts](https://github.com/DevOtts)._

_Part of the build-it plugin · authored by [DevOtts](https://github.com/DevOtts)._

_Part of the build-it plugin · authored by [DevOtts](https://github.com/DevOtts)._

_Part of the build-it plugin · authored by [DevOtts](https://github.com/DevOtts)._

--- _Authored by [DevOtts](https://github.com/DevOtts)._

--- _Authored by [DevOtts](https://github.com/DevOtts)._

--- _Authored by [DevOtts](https://github.com/DevOtts)._

--- _Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

_Authored by [DevOtts](https://github.com/DevOtts)._

Executable file

Executable file

Executable file

#!/usr/bin/env bash

#!/usr/bin/env bash

#!/usr/bin/env bash

#!/usr/bin/env bash

#!/usr/bin/env bash

#!/usr/bin/env bash

#!/usr/bin/env bash

#!/usr/bin/env bash

#!/usr/bin/env bash

Executable file

// ST-08 adversarial regression tests: every attack that got a fake done/verified/100%, a hidden blocked // task, or a silent lowering past the writer or the validator before ST-08, plus the defenses that held. // Authored by DevOtts (https

Executable file

// Authored by DevOtts (https://github.com/DevOtts).

Executable file

function exec(cmd, vars, cwd) {

const r = exec(cmds[name], varsFor(fx, id, extra), fx.root);

const rep = exec(cmds.report, varsFor(fx, 'T-1'), fx.root);

assert.equal(exec(cmds.start, varsFor(fx, 'T-1'), fx.root).code, 0);

const r = exec(cmds.finish, varsFor(fx, 'T-1', { '<evidence>': ev }), fx.root);

assert.equal(exec(cmds.start, varsFor(fx, 'T-1'), fx.root).code, 0);

let r = exec(noFlag, varsFor(fx, 'T-1'), fx.root);

r = exec(cmds.finish, varsFor(fx, 'T-1', { '<evidence>': '.taskstate/results/missing.md' }), fx.root);

r = exec(cmds.finish, varsFor(fx, 'T-1', { '<evidence>': '.taskstate/results/empty.md' }), fx.root);

Executable file

Executable file

Executable file

Executable file

test('schema file parses and carries the forbidden-key and conditional rules', () => {   const schema = JSON.parse(fs.readFileSync(path.join(path.dirname(LIB), 'status.schema.json'), 'utf8'));   assert.equal(schema.$schema, 'https://json-sc

const schema = JSON.parse(fs.readFileSync(path.join(path.dirname(LIB), 'status.schema.json'), 'utf8'));   assert.equal(schema.$schema, 'https://json-schema.org/draft/2020-12/schema');   assert.deepEqual(schema.$defs.forbidden.anyOf.map((x) 

assert.equal(schema.$schema, 'https://json-schema.org/draft/2020-12/schema');

Executable file

#!/usr/bin/env node

//     --lens <lens STATUS.md> --out <dir> [--verdict <review-it checkpoint report>] \ //     [--verdict-plain <review-it report without checkpoint lines>] [--only CASE,CASE] // Exit 0 when every case it ran passed, 1 otherwise. Authored by

// Exit 0 when every case it ran passed, 1 otherwise. Authored by DevOtts (https://github.com/DevOtts).

// R-1: re-run the Test Contract check on demand. The command comes from the plan committed at HEAD, never from // status.json. Authored by DevOtts (https://github.com/DevOtts).

Executable file

// status.json. Authored by DevOtts (https://github.com/DevOtts).

Executable file

Executable file

// init-it support in status-write: `session add --create` and role update on re-registration. // Authored by DevOtts (https://github.com/DevOtts).

Executable file

// Authored by DevOtts (https://github.com/DevOtts).

Executable file