Stats
Actions
Tags
From done-or-not
Discover Git context, project semantics, candidate features, and unknowns before any authenticity judgment.
How this command is triggered — by the user, by Claude, or both
Slash command
/done-or-not:discover-project-semanticsThe summary Claude sees in its command listing — used to decide when to auto-load this command
# discover-project-semantics ## When To Use This Command Use this command when you need to reconstruct project intent before judging implementation. This command is read-only and should be the default starting point for any serious audit. ## Operator Procedure 1. Inspect Git context before reading implementation details. 2. Detect whether the repository is a new repo, cold start, existing audited project, feature branch, dev baseline, or main baseline. 3. Read documents before code. 4. Inspect the repository shape and implementation surface conservatively. 5. Build a candidate featur...
Use this command when you need to reconstruct project intent before judging implementation. This command is read-only and should be the default starting point for any serious audit.
Implemented, Real, Mostly real, Fake, Non-operational, or
Drifted.Authenticity Status must default
to Not audited.docs/CLAUDE.md / AGENTS.mdInspect:
docs/implementation-audit/ already existsInterpret strictness conservatively:
main / master: stable baselinedev / develop: integration baselinefeature/*: active feature scopefix/* / bugfix/*: bug path scopeprototype/* / spike/*: placeholder-tolerant, but still evidence-awareIf the repo is new or nearly empty:
If the repo has code but no stable semantic audit context:
# Project Semantics Discovery
## Mode
- Mode: <selected mode>
- Reason:
## Git Context
- Current branch
- Working tree status
- Recent commits
- Changed files
- Likely base branch
- Branch interpretation
- Audit strictness implication
- Open PRs: Not checked from local context
- Remote: Not confirmed
## Documentation Sources
| Source | Found | Notes |
|---|---:|---|
## Repository Shape
| Area | Evidence | Interpretation |
|---|---|---|
## Candidate Feature Map
| ID | Feature | Presence Status | Authenticity Status | Evidence | Audit Needed / Notes |
|---|---|---|---|---|---|
## Unknowns
1.
2.
3.
## Recommended Next Step
- These unknowns are recommended entry points for scoped feature audits.
- You do not need to resolve all unknowns before auditing one feature.
- Pick one unknown or one feature and run `/done-or-not:audit-feature`.
## Audit Memory Status
- `docs/implementation-audit/`: Found / Not found
- Files written by this command: None
- Reason: discovery is read-only
## To Save This Discovery
Run:
1. `/done-or-not:propose-audit-docs`
2. Review the proposed docs
3. `/done-or-not:save-audit-docs`
Suggested append-only record path:
`docs/implementation-audit/audits/YYYY-MM-DDTHHMM-project-semantics-discovery.md`
The mode field must print exactly one selected mode, not a mixed menu. A short explanation may follow the selected mode.
npx claudepluginhub fofxjc/done-or-not --plugin done-or-not