By FoFxjc
Check whether AI-assisted feature work is actually done by auditing implementation evidence.
Audit a confirmed feature through an implementation evidence chain without modifying product code.
Audit whether the current branch is ready to merge, including placeholder carryover and documentation alignment.
Discover Git context, project semantics, candidate features, and unknowns before any authenticity judgment.
Propose draft implementation-audit docs for docs/implementation-audit without writing files.
Save approved audit-memory docs under docs/implementation-audit only.
Check whether "done" is actually done.
Done or Not is a Claude Code plugin for auditing whether claimed AI-assisted feature work is actually implemented.
It is markdown-first and audit-only: it reads a repository, compares
claims against implementation evidence, and can save audit memory under
docs/implementation-audit/ in the target repo.
Current version: v0.1.2
docs/implementation-audit/ with approvalFrom the CLI:
claude plugin marketplace add FoFxjc/done-or-not
claude plugin install done-or-not@fofxjc
From within a Claude Code session:
/plugin marketplace add FoFxjc/done-or-not
/plugin install done-or-not@fofxjc
From the CLI:
claude plugin marketplace update fofxjc
claude plugin update done-or-not@fofxjc
From within a Claude Code session:
/plugin marketplace update fofxjc
/plugin update done-or-not@fofxjc
Primary commands:
/done-or-not:discover-project-semantics
/done-or-not:propose-audit-docs
/done-or-not:save-audit-docs
/done-or-not:audit-feature
/done-or-not:audit-merge-readiness
Typical flow:
Discover project semantics
-> propose audit docs
-> save audit baseline
-> audit one feature
-> audit merge readiness when needed
Done or Not is a lightweight single-repo completion legitimacy layer for AI-assisted feature work.
It validates whether claimed feature and project implementations inside a repository are actually real.
It does not call a feature fake just because a code path looks suspicious. It first establishes project semantics, then allows stronger authenticity judgments only after scope is confirmed or strongly evidenced.
Core rule:
No semantics -> no fake verdict.
Draft semantics -> preliminary audit only.
Confirmed semantics -> real authenticity gate.
docs/implementation-audit/
├── README.md
├── feature-map.md
├── implementation-semantics.md
├── placeholder-policy.md
├── git-context.md
├── audits/
├── decisions/
└── merges/
This folder is the project audit trail. It does not replace official product docs.
Done or Not uses a split status model:
| Model Part | Allowed Values |
|---|---|
| Presence Status | Claimed, Observed, Inferred, Confirmed |
| Authenticity Status | Not audited, Preliminary, Partial, Mostly real, Real, Non-operational, Fake, Drifted |
Discovery and proposal phases never use final verdict language. During
discovery, features keep Authenticity Status: Not audited.
Validate the plugin manifest:
claude plugin validate .claude-plugin/plugin.json
Validate the marketplace manifest:
claude plugin validate .claude-plugin/marketplace.json
Load the plugin directly for local dogfooding:
claude --plugin-dir /path/to/done-or-not
If /plugin opens the generic plugin management UI and does not clearly
list commands in your terminal, inspect the commands/
directory or run a known namespaced command directly.
If claude --plugin-dir fails with an EPERM error involving
~/.claude.json, that is a local Claude CLI config or permission
issue, not a plugin validation failure. Some sandboxed runners also
cannot write under ~/.claude*, which can block local plugin loading
even when the plugin itself is valid.
done-or-not/
├── .claude-plugin/
│ ├── plugin.json
│ └── marketplace.json
├── commands/
├── skills/
├── templates/
├── docs/
│ ├── positioning.md
│ └── roadmap.md
├── CLAUDE.md
├── README.md
└── LICENSE
Done or Not is a lightweight single-repo completion legitimacy layer for AI-assisted feature work.
It validates whether claimed feature and project implementations inside a repository are actually real.
It is not an Agent OS, orchestration engine, workflow platform, memory system, task tracker, or multi-agent controller.
It focuses on repo-local feature claims, implementation evidence, authenticity verdicts, and audit records.
See docs/roadmap.md for the path from the current plugin to v1.0. The roadmap is the source of truth for milestone scope and non-goals.
Apache-2.0
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub fofxjc/done-or-not --plugin done-or-notOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimComprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Standalone image generation plugin using Nano Banana MCP server. Generates and edits images, icons, diagrams, patterns, and visual assets via Gemini image models. No Gemini CLI dependency required.
Multi-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
Write feature specs, plan roadmaps, and synthesize user research faster. Keep stakeholders updated and stay ahead of the competitive landscape.