Install
$ agentstack add skill-jasontang-ai-acp-deliberation-cycle ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
What this skill is for
Use this skill when the task is about the whole ACP cycle: creation, submission, routing when applicable, release, reflection, archive, and audit handoff.
When to use it
- A user asks to run a complete demo or pilot cycle.
- A cycle is stuck and the next valid lifecycle action must be identified.
- A reviewer needs to understand the end-to-end ACP workflow.
When not to use it
- Use
epistemic-routingfor routing-only inspection. - Use
digest-and-explanationfor digest-only review. - Use
pilot-analysisfor metric comparison after cycles are complete.
Inputs expected
- Cycle condition:
interventionorbaseline_thread - Title and prompt
- Participant roster or participant seed file
- Current cycle status
- Desired next operator action
Preflight / prerequisites
- Confirm the canonical lifecycle in
docs/specs/ACP_PROTOCOL_CONTRACT_SPEC.md. - Build the repo before CLI execution when
dist/may be stale. - Confirm whether the task is review-only or allowed to mutate cycle state.
- For intervention cycles, confirm submissions are closed before routing.
Workflow
- Inspect the cycle state and condition.
- Identify the next valid lifecycle transition.
- For
intervention, route before digest release. - For
baseline_thread, keep output chronological and do not invent routing evidence. - Run audit/export review before claiming the cycle is demonstration-ready.
- Archive only after reflection or review requirements are satisfied.
Decision rules / judgment criteria
- Lifecycle validity beats operator convenience.
- Baseline and intervention conditions must remain comparable.
- Never convert a baseline thread into routed evidence.
- Missing audit/export evidence blocks claims, even when the UI looks correct.
Escalation rules
Escalate to a maintainer when the requested action would skip a lifecycle state, erase condition comparability, or require interpreting missing participant data.
Available scripts
scripts/check-lifecycle.mjs: inspects a cycle JSON file or demo lifecycle seed for canonical status vocabulary and next-action notes.
Outputs
- Next valid lifecycle action
- Cycle readiness notes
- Audit-ready handoff summary
Failure handling
Stop on invalid transitions. Report the current status, requested action, and the canonical transition that must happen first.
Trust / safety notes
This skill can guide lifecycle-changing commands. Treat it as operator-sensitive and do not mutate a cycle unless the user explicitly requested execution.
Composition notes
Use this before pilot-cycle-review when the cycle state is uncertain. Use operator-audit after lifecycle actions to prove what happened.
Examples to inspect next
Read examples/casebook.md for happy path, failure path, ambiguity, and anti-pattern examples.
Evaluation hooks
Run scripts/check-lifecycle.mjs or npm run skills:audit to verify this skill's structure and lifecycle vocabulary.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: jasontang-ai
- Source: jasontang-ai/acp
- License: Apache-2.0
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.