| name | aca |
| description | Adaptive certificate-driven reasoning for non-trivial engineering tasks in this repo, including semantic PR review, patch equivalence checks, debugging, safe implementation, refactors, codebase QA, and static risk review. Use when the task is cross-file, high-risk, behaviorally sensitive, or execution-constrained. Do not use for trivial mechanical edits or obvious single-file changes. |
You are running the ACA skill.
Treat everything passed with $aca as the task request.
Immediate actions
- Parse the request. If the first token is a mode, use it:
- review
- verify_patch
- debug
- implement
- qa
- risk_review
- refactor
- test_design
- If the mode is not explicit, infer it from the task.
- For non-trivial tasks, load and follow:
references/reference.md
assets/certificate-template.md
ACA operating policy
Adapt the certificate to the repo and task before deep work.
Required sequence
- Normalize the task and success condition.
- Calibrate the repo.
- Synthesize the adapted certificate plan.
- Use a hypothesis-driven inspection loop.
- Separate VERIFIED, INFERRED, and UNKNOWN.
- Try to falsify the leading answer.
- Return the result with a concise but auditable certificate.
Mode heads
review
- summarize semantic delta
- identify externally visible behavior changes
- inspect off-diff impact
- identify missing coverage and merge risk
verify_patch
- define equivalence in repo terms
- compare behavior slices
- find a concrete counterexample if non-equivalent
- end with YES or NO
debug
- separate symptom site from cause site
- rank candidate fault locations
- trace the divergence path
- propose the smallest credible fix direction
implement
- define intended behavioral delta
- preserve repo invariants
- prefer the minimal sufficient change set
- identify tests/docs/contracts to update
qa
- answer directly, but only after tracing the relevant semantics
- state what would need to be true for the opposite answer
risk_review
- inspect trust boundaries, unsafe defaults, validation gaps, auth drift, and failure amplification
refactor
- define the refactoring goal (readability, coupling reduction, performance, testability, etc.)
- preserve all externally visible behavior
- identify callers and dependents that must remain unchanged
- prefer incremental, reviewable steps over large rewrites
test_design
- identify the behavior contract to be tested
- derive test cases from the code's decision points and boundary conditions
- cover happy paths, edge cases, error paths, and state transitions
- specify what each test proves about system correctness
Final output
Use this structure unless the user explicitly requests another format:
- Task Understanding
- Repo Calibration
- Adapted Certificate Plan
- Findings
- Verdict
- Certificate
- Risks / Gaps
- Confidence
- Smallest Next Step