About grace-execute
A workflow skill for executing projects governed by approved GRACE spec and plan bundles. It enforces baseline and final lint checks, task verification, controlled parallel work, and explicit confirmation before applying and archiving a bundle.
Key features
- 01Approved bundle and identity checks
- 02Baseline and final lint gates
- 03Task-level acceptance verification
- 04Parallel-safety preflight
- 05Durable-state reconciliation
- 06Explicit apply confirmation
Use cases
- 01Executing an approved GRACE project plan
- 02Verifying work before applying a bundle
- 03Stopping safely when project drift is unexplained
Install
Copy the command and run it in a terminal.
npx skills add https:/ / github.com/ osovv/ grace-marketplace --skill grace-executeAfter installing, type/grace-executein Claude Code to confirm it loaded.
Bundled filesagents/openai.yaml
Details
From the project README<assertion_commands>
- Active-baseline preflight before observed writes:
grace lint --path PROJECT --assertions current - Selected baseline:
grace lint --path PROJECT --change C-ID --assertions baseline(add--run-commandswhen the baseline declaresMustPassCommand) - Selected target without commands:
grace lint --path PROJECT --change C-ID --assertions target - Selected target with command evidence:
grace lint --path PROJECT --change C-ID --assertions target --run-commands - Final end-state validation:
grace lint --path PROJECT --change C-ID --assertions final(add--run-commandswhen the target declaresMustPassCommand) - Parallel preflight:
grace lint --path PROJECT --parallel-preflight</assertion_commands>
<mode_selection>
Wait for explicit sequential or parallel-safe choice. Parallel-safe requires the explicit preflight to pass. Workers never mutate approved plans; durable .grace changes are applied centrally after observed work verifies.
</mode_selection>
<recovery_decision_table> | state | required action | | clean-to-start | Run selected baseline, then execute tasks. | | partial-observed-writes | Inspect the declared observed scope and ask whether to resume or revert. | | durable-state-changed | Hard stop; supersede and replan. Approved assertions are immutable. | | target-already-satisfied | Run final end-state validation, opted-in command evidence when declared, durable reconciliation, and ask for explicit apply confirmation. | | unsafe-unknown-drift | Hard stop and report unexplained files. | </recovery_decision_table>
<execution_rules>
- Run the selected baseline before implementation, including explicit
--run-commandswhen its assertions declareMustPassCommand. - Execute one dependency-ready task or one verified parallel-safe batch at a time.
- Run each task's acceptance and verification immediately.
- Apply approved durable context, graph, and verification changes centrally.
- Reconcile durable state, run leaf plan gates, then run selected
--assertions finalas the outermost lifecycle gate, including--run-commandswhenMustPassCommandis declared. Final mode performs full project lint, evaluates the selected target, keeps unrelated approved baselines active, and does not re-evaluate the selected plan's superseded baseline. - Ask for explicit apply confirmation after fresh end-state evidence passes.
- Only then set spec and plan to
appliedand archive the complete bundle. - Never edit approved assertions/scopes/tasks in place, bypass stale evidence, or continue through unknown drift. </execution_rules>