Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/jeonje/claude-plugins/manpx skills add JeonJe/claude-plugins --skill magit clone --depth 1 https://github.com/JeonJe/claude-pluginsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/jeonje/claude-plugins/ma)<a href="https://agentmods.dev/skills/jeonje/claude-plugins/ma"><img src="https://agentmods.dev/badge/skills/jeonje/claude-plugins/ma.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00061 | $0.02440 |
| Opus 5 | $0.00030 | $0.01220 |
| Sonnet 5 | $0.00012 | $0.00488 |
| Haiku 4.5 | $0.00006 | $0.00244 |
Grade A, and why
ma scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MA quality-gated pipeline
Coordinate call order, scope, state, approval, validation, and (optional) notification. Keep phase implementation in the corresponding ma-* skill; do not duplicate it here.
All paths in this skill resolve through the workspace config (see the plugin's config/README.md): artifacts under {contextRoot}/{domain}/, pipeline state under {stateRoot}/{key}/.
Invocation
/ma <issue-key> [--resume] [--base <branch>] [--target <branch>] [--auto]
- Require an issue key matching the configured
issueKeyPattern, e.g.PROJ-123. If the workspace has no issue tracker, any key matching the pattern names the work; requirements then arrive as user-provided text. - Default
--baseto the project's configured base branch (commonlydevelopormain). - Default
--targettofeature/{key}-{slug}. - Use
--resumeto continue from the first incomplete phase after validating recorded artifacts. - Treat
--autoas flag forwarding only. It does not grant implementation approval unless the user explicitly requested unattended implementation.
Examples:
| Input | Action | Stop condition |
|---|---|---|
/ma PROJ-123 |
Resolve the requested scope, initialize state, then run the gated pipeline | Selected scope terminal |
/ma PROJ-123 PLAN까지만 |
Run PRD, design, and per-project PLAN | Phase 2.3 |
/ma PROJ-123 --resume |
Revalidate prior results and enter the first incomplete phase | Selected scope terminal or a halt gate |
References index
| File | When to read |
|---|---|
| references/scope-gate.md | Phase 0 and every MA entry — terminal scope resolution (SSOT, also read by each ma-* skill) |
| references/accuracy.md | At the start of every ma-* step — accuracy and reporting contract |
Execution contract
- Resolve scope before work. Read
references/scope-gate.mdand set one ofdocs,plan,plan-review, orimpl. Interview once when the terminal scope is materially ambiguous. Never enter Phase 3+ for scopes belowimpl. - Initialize durable state immediately after scope resolution. This preserves recovery if any later phase fails.
- Enter phases in order. Before each phase, confirm the preceding phase is in
completedPhasesand its required artifact exists. If handoff moved an artifact, validate the new path recorded in state instead of requiring the old path. - Require implementation approval. After Phase 2.5, stop once for approval that records project, base, and target. Neither
--autonor animplscope bypasses this gate unless unattended implementation approval was explicit. - Keep project writes inside the handoff worktree. Phase 4 implementation goes through
/ma-dispatch; do not edit a target project's main checkout directly. - Complete each phase atomically. Call the phase skill, verify its artifact gate, send the optional notification, append
timeline.md, updatephaseResults, then add the phase tocompletedPhases. A failed step leaves the phase incomplete and must be repaired and rechecked. - Use agent return values, not polling. The
/ma-dispatchSTATUS block is the completion source of truth. Do not addsleeploops. - Preserve recoverability on halt. Record structured evidence, send the optional halt notification, and keep artifacts intact so
--resumecan continue safely.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 142 lines · 61 tokens per session scan A 93cdf8d4bc2c
ma is a skill published in the GitHub repository JeonJe/claude-plugins (3 stars, last pushed 1mo ago), licensed MIT. It adds 61 tokens to every session and 2,440 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
blueprint-feature-tracker-sync
Sync feature tracker with TODO.md, taskwarrior sidecars, and PRDs. Use when reconciling TODO.md vs tracker, draining WO entries, or recalculating stats.
blueprint-status
Show blueprint version, config, PRD/ADR/PRP counts, and feature tracker progress. Use when auditing traceability, orphan docs, or stale generated content.
comfy-node
Orchestrate a ComfyUI node pack from idea to registry: scaffold, create + seed the repo, open the gitops adoption PR. Use when releasing or spinning up a new comfyui node pack.
blueprint-work-order
Create a work-order for isolated subagent execution, optionally linked to a GitHub issue. Use when breaking a PRP into delegatable tasks or spawning from an issue.
blueprint-autonomy-level3
Install blueprint autonomy level 3 (ADR-0020): scheduled-autorun + approved-work-order-execution GitHub workflows. Use when enabling out-of-band blueprint automation in a repo.
blueprint-execute
Blueprint meta command: determine and execute the next logical action. Use when asked 'what's next?', 'continue blueprint', or 'run blueprint' without a subcommand.