agent-done-check — Commit Acceptance Verifier
Map explicit acceptance criteria to commit-bound checks and report passed, failed or unverified with traceable evidence.
$1
Works with the AI Agents you already use
agent-done-check — Commit Acceptance Verifier
Example session with this skill installed
Verify that a fixed trusted commit contains an MIT license and can run Node. Validate config first, run the two checks, then verify the evidence bundle.
- Read your context and instructions
- Compiled the agent-done-check
The synthetic config validated with criterionCount=2 and checkCount=2. Against source commit cd3e1451c70bd59ce9c03e1b3a768a42653a1f9b, agent-done-check 0.45.0 reported overall status passed. The license-check read the committed LICENSE and matched the MIT License substring; runtime-check executed the argv command [node,--version] and matched its expected stdout substring. The report recorded the full commit and config hash. Bundle verification returned valid=true, artifactsChecked=3, errors=[]. This narrow example does not prove application behavior or requirement completeness. For a real coding task, replace these demonstration criteria with meaningful behavioral assertions, keep unverified coverage explicit, and inspect the exact target revision before making a completion claim.
Connects securely to your tools. The creator never sees your data.
About this skill
Original agent workflows for verifying a coding task against an exact Git commit: map requirements to observable checks, review executable policy, validate config without executing it, run commit-bound checks, and interpret reports and manifest consistency.
Requires agent-done-check 0.45.0, Node.js >=22 and Git. Command/file checks work without a model API key or browser runtime. Playwright is optional and separately installed. The runtime is experimental/pre-1.0.
Verified pinned baseline: 46 tests passed. A synthetic MIT-license and Node-runtime acceptance config validated, passed both criteria against a fixed commit, and its three-artifact bundle verified successfully. This establishes only those narrow configured assertions, not general application completion. Production dependency audit reported zero advisories for the tested source lockfile.
Detached worktrees isolate source revisions, not host privileges. Checks are not sandboxed; trusted repositories/configs only. Uncommitted edits are outside commit-bound coverage. Keep failed and unverified distinct.
The MIT runtime remains free at https://github.com/shivam039/agent-done-check. This original MIT workflow package adds curation and examples, not exclusive runtime access.
Pre-1.0 runtime. A pass proves only configured assertions. Detached worktrees are not sandboxes; uncommitted edits are not verified. Missing setup/commit binding and omitted checks remain unverified. Bundle hashes do not prove authorship.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 19 days ago
Requires
Permissions
Allowed Hosts
Install only in an authorized project; review commands/config and endpoints before execution. Read task-authorized files and write only authorized project changes/reports. Runtime and source packages remain free and open-source under MIT; this purchase is for original curated instructions.