WenyuChiou/agent-collab-skills

# agent-acceptance-gate

The read-only check between "the lanes are done" and "commit or push". It
reconciles results, runs the triggered presets, and returns a verdict. It
never merges, commits, retries, or turns a decline, cancellation, timeout, or
failure into success. Since 0.6.0 it includes the former
`agent-output-reconciler`.

## Inputs

- the lane table or plan: lane ids, roles, scopes, acceptance criteria;
- the job's stated outcome and outside check, if a plan header exists (for
  example from the `aim-higher` skill): the merged result is checked against
  them, not only against each lane's own criteria;

# agent-context-budget

Keep the primary session small while lanes do the reading and writing. There
is no second budget source: native mode follows the host's limits and the caps
written into each brief; harness mode takes every number from the policy
referenced by the current `TaskCheckpoint`. The "roughly 2 KB" in the
description says when to use this skill; it is not a limit.

Do not use this skill to summarize away unresolved failures, human decisions,
raw scientific evidence, or security findings.

## Before launching lanes

- Put a return cap in every brief: status, short summary, files changed,

# agent-debate

A bounded adversarial review of one consequential trade-off. Debate makes the
evidence visible to the decision maker; it does not settle truth by vote and
does not authorize implementation.

## Use only when

- both options are plausible and materially different;
- the decision is consequential enough to justify two independent reviewers;
- identifiable evidence could inform or falsify each side.

Do not debate a factual question that a direct check can settle, a naming
preference, or a decision an authorized human already made.

# agent-output-reconciler (redirect)

Merged into [agent-acceptance-gate](../agent-acceptance-gate/SKILL.md) in
0.6.0. Run that skill: its "Step 1: Reconcile" is the former workflow, and the
report template moved to
[reconciliation_template.md](../agent-acceptance-gate/references/reconciliation_template.md).

A plan that hands a round to the reconciler and then to the gate now needs one
gate pass. The rules are unchanged: missing and failed results stay distinct,
scope drift is recorded, and a reconciliation is not an acceptance decision.

This stub is removed in 0.7.0. See [migration 0.6](../../docs/migration-0.6.md).

# agent-plan-act-reflect (deprecated)

Deprecated in 0.6.0 and removed in 0.7.0.

- Iterating with the user present (edit, run the check, fix, repeat) needs no
  skill. Keep the acceptance check runnable and report its real result.
- An unattended multi-cycle loop under a configured `agent-collab` policy
  follows [harness-loop.md](../agent-context-budget/references/harness-loop.md),
  the former body of this skill. It still runs `agent-collab policy evaluate`
  after every cycle and before every spawn.
- Comparing two consequential alternatives is
  [agent-debate](../agent-debate/SKILL.md), not a loop.

See [migration 0.6](../../docs/migration-0.6.md).

# agent-shared-memory (redirect)

Merged into [agent-context-budget](../agent-context-budget/SKILL.md) in 0.6.0.
Use its "Durable decisions" section. The event and proposal schema, approval
binding, and read/propose/decide/apply modes moved to
[coord_memory_schema.md](../agent-context-budget/references/coord_memory_schema.md).

The contract is unchanged: memory output is a proposal, only an explicit human
approval lets it apply, and applying appends a new event that never edits an
existing one.

This stub is removed in 0.7.0. See [migration 0.6](../../docs/migration-0.6.md).

# agent-task-splitter (fan-out planner)

Plan a fan-out before launching it. This skill plans; the host launches the
lanes. The name stays `agent-task-splitter` so existing prompts and plans keep
working.

## Pick the mode

| Mode | Use when | Artifacts |
|---|---|---|
| Native (default) | All lanes run inside one host session: parallel or background subagents, workflow lanes, one-shot CLI delegates, with the user present | Lane table in the conversation and one brief per lane. No `.coord/` plan, packets, or checkpoint. |
| Harness | A round spans sessions or tools, an unattended loop runs, work must resume after an interruption, a signed human gate is needed, or `AGENT_COLLAB_POLICY` is configured | `.coord/plan.yml` (schema v2) and `.ai/task_<NNN>_<slug>.md` packets; see [references/harness-plan.md](references/harness-plan.md) |

A configured policy always selects harness mode, and an unreadable configured