agent-acceptance-gate
Use when reconciled multi-agent output needs an evidence-based PASS, CONDITIONAL PASS, or FAIL record before authorized delivery.
Use when a user asks to inspect coordination memory, propose a durable decision or resolution, review pending memory proposals, or apply an explicitly approved proposal as a new append-only event.
$ npx -y skills add WenyuChiou/agent-collab-skills --skill agent-shared-memory --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/agent-shared-memoryContext preview
The summary Claude sees to decide when to auto-load this skill.
Use when a user asks to inspect coordination memory, propose a durable decision or resolution, review pending memory proposals, or apply an explicitly approved proposal as a new append-only event.
name: agent-shared-memory description: Use when a user asks to inspect coordination memory, propose a durable decision or resolution, review pending memory proposals, or apply an explicitly approved proposal as a new append-only event.
Coordinate durable facts without letting an agent silently rewrite project memory. Read the event schema in references/coord_memory_schema.md before creating or applying a proposal.
Repository state, current evidence, and recorded human decisions outrank memory. Recall systems are optional caches.
All memory mutations start as proposals:
.coord/memory-proposals/<proposal-id>.json
Without an explicit human approve decision, do not add, supersede, archive, delete, overwrite, or compact canonical memory.
Read canonical events and return:
Do not load raw agent logs into the primary session. Follow their stable paths only when needed.
Create a proposal with:
{ "schema_version": 1, "proposal_id": "<uuid>", "source_task": "<task id>", "action": { "operation": "add | supersede | archive | delete", "target_ref": null, "summary": "<compact proposed event>", "evidence_refs": ["<stable ref>"] }, "state": "proposed", "created_at": "<ISO 8601>", "decision": null }
Supersede, archive, and delete require a specific `action.target_ref`. A proposal with no `action.evidence_refs` remains pending and must not be applied. The `action` object is immutable. Its canonical JSON bytes, not the mutable proposal envelope, are the approval payload.
Only a human may approve, decline, or revise a proposal. Record:
Decline and timeout remain non-success. A revise decision creates or updates a proposal; it is not approval.
Apply only a proposal whose recorded state is approved and whose immutable `action` bytes still match `decision.affected_action_hash`. Decision metadata and state are outside that hashed payload, so recording approval does not invalidate the approval hash. The decision uses the same `hmac-sha256` authorization boundary as checkpoint human records. Do not expose the signing secret to a delegated executor.
Application appends one immutable event to .coord/memory.yml. It does not edit the target event:
It never edits an existing event; a resolution, correction, or supersession is a new append-only event.
repository-specific retention decision.
Use this memory for agent coordination decisions, questions, artifact pointers, and execution outcomes. Research truth belongs in research evidence stores; paper claims belong in the paper's claim/evidence contract.
.coord/ is scratch and gitignored by default. Do not commit every proposal or agent boundary. Promote only an explicitly selected checkpoint snapshot, shipping artifact, or acceptance record into a repository-owned evidence path.
Always report:
Never report an unapproved proposal as a canonical decision.
繁體中文 · Public harness contract · 0.4 migration · 0.5 goal-slice migration Provider-neutral collaboration skills plus an optional, standard-library-first Python harness for bounded, resumable, human-authorized agent work.
Use when reconciled multi-agent output needs an evidence-based PASS, CONDITIONAL PASS, or FAIL record before authorized delivery.
Use when an agent task needs bounded context packets, checkpoint-aware compaction, a fresh-session primer, or policy-controlled child results without loading…
Use when a consequential decision has genuinely competing positions that need bounded adversarial review before a human decision.
Use when multiple agent results must be checked for missing evidence, scope drift, contradictions, and merge conflicts before an acceptance decision.
Use when one agent must iteratively plan, act, evaluate evidence, and revise under a readable policy until acceptance, checkpoint, stop, or human escalation.
Use when a goal needs multiple bounded agent roles, parallel or dependent task packets, explicit scope partitions, and a provider-neutral DAG. Not for…