Plugin MaintenanceSAFE
The Agent Harness for AI-Human Collaboration, inspired by the AI-DLC (AI-Driven Development Lifecycle)
Overview
The Agent Harness for AI-Human Collaboration, inspired by the AI-DLC (AI-Driven Development Lifecycle)
3bb5b08404b5OBSERVED · 2026-10-12Host compatibility
What the documentation claims. We have not run a compatibility test.
| Host | Status | Notes |
|---|---|---|
| claude-code | mentioned | |
| codex | mentioned | |
| cursor | mentioned | |
| openclaw | mentioned |
What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
---
name: plugin-maintenance
description: Guide for modifying the Chorus plugin (Claude Code, Codex, OpenClaw, Kiro, Pi, dsh, and Hermes ports), updating skill documentation, and releasing new plugin versions.
license: AGPL-3.0
metadata:
author: chorus
version: "0.5.0"
category: development
---
# Chorus Plugin & Skill Maintenance
How to modify the Chorus plugin, update skill documentation, and release new versions. **Seven plugin packages** are maintained in parallel — Claude Code, Codex, OpenClaw, Kiro, Pi, dsh (DeepSeek Harness), and Hermes Agent — plus the standalone skill surface. That is **eight skill surfaces total**; when you change skill content, sweep all eight (see [Skill Content Changes — Seven Surfaces](#skill-content-changes--seven-surfaces), which also lists the Hermes copy). The Hermes port has its own section: [Hermes Agent port](#hermes-agent-port).
## File Structure
```
.claude-plugin/
marketplace.json ← Claude Code marketplace registry (version here)
public/chorus-plugin/ ← Claude Code plugin package
.claude-plugin/
plugin.json ← Plugin metadata (version here)
hooks.json ← Hook definitions (SubagentStart, etc.)
bin/ ← Hook scripts (bash) — stateful via API state-get/set
skills/
chorus/SKILL.md ← Core skill
develop/ idea/ proposal/ quick-dev/ review/ yolo/SKILL.md
agents/ ← Reviewer agents as .md (Claude Code style)
proposal-reviewer.md
task-reviewer.md
plugins/chorus/ ← Codex plugin package (separate from Claude Code)
.codex-plugin/
plugin.json ← Plugin metadata (version here)
hooks.json
hooks/ ← Hook scripts (bash) — intentionally stateless
on-session-start.sh
on-post-submit-proposal.sh
on-post-submit-for-verify.sh
chorus-mcp-call.sh ← MCP helper (has hardcoded clientInfo version)
hook-output.sh
skills/
chorus/SKILL.md ← Codex port — mentions $skill syntax, ~/.codex/config.toml
develop/ idea/ proposal/ quick-dev/ review/ yolo/SKILL.md
chorus-proposal-reviewer/SKILL.md ← Reviewers as skills in Codex (no agents/)
chorus-task-reviewer/SKILL.md
packages/openclaw-plugin/ ← OpenClaw plugin package (TS runtime + skills)
openclaw.plugin.json ← Plugin manifest (id, skills dir, activation, configSchema) — NO version field
package.json ← npm package (version here; `openclaw` block: extensions, runtimeExtensions, install/compat)
src/ ← TypeScript runtime (index.ts, mcp-client.ts, sse-listener.ts, event-router.ts, wake.ts)
— real-time SSE event bridge + MCP registration; NOT bash hooks
dist/ ← Compiled JS (npm install loads this; linked install loads src/ via jiti)
skills/
chorus/SKILL.md ← OpenClaw port — tools namespaced `chorus__<tool>`, inline OpenSpec detection
develop/ idea/ proposal/ quick-dev/ review/ yolo/ brainstorm/ openspec-aware/SKILL.md
proposal-reviewer/SKILL.md ← Reviewers as skills (like Codex, no agents/)
task-reviewer/SKILL.md
public/kiro-plugin/ ← Kiro CLI plugin (loose .kiro/ template tree + install script)
.kiro/
settings/mcp.json ← Chorus remote MCP server (${env:CHORUS_API_KEY} bearer, disabled:false)
skills/chorus-*/SKILL.md ← the chorus-PREFIXED skills (no bare names — global-install distinctiveness)
agents/chorus.json ← main agent (.json — Kiro CLI, NOT .md); hosts all hooks via __CHORUS_BIN__ placeholder
agents/chorus.md ← main-agent system-prompt sidecar (file://./chorus.md)
agents/chorus-*-reviewer.json ← 3 read-only reviewer subagents (tools:["read","@chorus"])
steering/chorus.md ← platform overview + AI-DLC context (folds in the `chorus` overview skill)
bin/ ← Hook scripts (bash, 3.2-safe) + chorus-api.sh + test-syntax.sh
installer copies these into <KIRO_DIR>/chorus-bin/ and resolves __CHORUS_BIN__
(public/install-kiro.sh) ← deprecation stub → redirects to `chorus agents add`; the .kiro/ tree is installed by cli/init/file-template.mjs (asset list: public/kiro-plugin/manifest.txt)
packages/chorus-pi/ ← Pi coding agent package (TS extension + skills)
package.json ← npm package + Pi manifest (version here)
extensions/chorus.ts ← Native Pi event handlers; auto checkin/session lifecycle/reviewer nudges
lib/lib.ts ← Pure helpers used by the extension and unit tests
bin/chorus-mcp-call.sh ← MCP helper for byte-exact OpenSpec document mirroring
skills/
chorus/SKILL.md ← Core skill; Pi `/skill:<name>` syntax
develop/ idea/ proposal/ quick-dev/ review/ yolo/ openspec-aware/SKILL.md
agents/ ← 3 read-only reviewer agents copied to ~/.pi/agent/agents/
test/ ← Static, helper-unit, extension-event, and manual-session verification
packages/chorus-dsh/ ← DeepSeek Harness (dsh) plugin — published npm bundle @chorus-aidlc/chorus-dsh
package.json ← npm package + dsh bundle patch (version here; tracks the APP version, e.g. 0.16.3)
src/index.ts ← Cordis plugin apply(): injects skills/persona/MCP wrapper; daemon-backend gate
cordis.patch.yml ← dsh composition patch shipped in the bundle
bin/chorus-mcp-call.mjs ← Node MCP-over-HTTP helper (clientInfo.version auto-read from package.json — no manual bump)
skills/
chorus/SKILL.md ← Core overview (dsh HAS a chorus/ overview, unlike Kiro)
<stage>-chorus/SKILL.md ← 13 stage/reviewer skills, `-chorus` SUFFIX (brainstorm/develop/idea/
proposal/quick-dev/review/yolo/docs/orchestTrust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | NA |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (1)
.agents/skills
Gates applied: no_behavioural_pass.
3bb5b08404b5full audit observations/trust-audit/skill/chorus-aidlc__plugin-maintenance.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-12 | 3bb5b08404b5 | SAFE | B | 89 | first audit |
Questions
What does the Plugin Maintenance skill do?
The Agent Harness for AI-Human Collaboration, inspired by the AI-DLC (AI-Driven Development Lifecycle)
Is Plugin Maintenance safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Plugin Maintenance access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
Which assistants does Plugin Maintenance work with?
Its documentation mentions claude-code, codex, cursor and openclaw. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (3bb5b08404b5), read on 2026-10-12. The repository is watched, and a new audit runs when it changes — this is the first audit.