Merge: agentic-harness v0.3.1 (P0 consistency & enforcement)

Author: MrBogomips

.claude-plugin/marketplace.json

@@ -14,7 +14,7 @@
       "name": "agentic-harness",
       "source": "./agentic-harness",
       "description": "Stand up, assess, and maintain an agentic harness in an existing repo — generate project-specific agent teams and the skills they use, then assess how effectively they are used",
-      "version": "0.3.0",
+      "version": "0.3.1",
       "category": "engineering",
       "tags": ["harness", "agents", "skills", "scaffolding", "orchestration", "multi-agent", "meta-skill"]
     },

CLAUDE.md

@@ -36,6 +36,12 @@ bash tests/ci/run-structural-tests.sh
 
 This is a strict requirement. No change is complete until validation passes. Requires `jq`.
 
+For a plugin headed to the community marketplace, also run the check the review pipeline runs before submitting:
+
+```bash
+claude plugin validate ./<plugin> --strict
+```
+
 ## CONNECTORS pattern
 
 Plugins with optional MCP server dependencies use `CONNECTORS.md` to declare `~~placeholder` categories. Skills reference these as `~~knowledge base`, `~~email`, etc. and degrade gracefully when no server is connected. Validation scripts check that every `~~placeholder` in SKILL.md files has a matching entry in CONNECTORS.md.

README.md

@@ -4,26 +4,39 @@ A curated collection of [Claude Code](https://claude.com/claude-code) plugins fo
 
 ## Plugins
 
-| Plugin | Description | Category | Version |
-|--------|-------------|----------|---------|
-| [agentic-harness](./agentic-harness) | Stand up, assess, and maintain an agentic harness — generate project-specific agent teams and the skills they use, then assess how effectively they are used | Engineering | 0.3.0 |
-| [developer-tools](./developer-tools) | Developer environment tooling — devcontainer generation, stack detection, infrastructure config | Engineering | 2.4.0 |
-| [human-resources](./human-resources) | HR interview workflow — job descriptions, pre-screening, interview prep, evaluation, compliance | Human Resources | 0.2.0 |
-| [kaizen](./kaizen) | Continuous improvement loops — recursive optimization engine with profiles for Claude Code usage, refactoring, and process improvement | Engineering | 1.0.0 |
-| [plantuml](./plantuml) | PlantUML diagrams — policy-driven authoring, rendering, lint, validate, review, advisor, and migrate | Documentation | 1.0.0 |
-| [project-management](./project-management) | SOW writing, review, estimation, and PMI-compliant PERT analysis — integrated project management pipeline | Operations | 1.0.0 |
-| [tech-writing](./tech-writing) | Technical writing support — documentation structure, style guides, content review | Documentation | 0.1.0 |
+| Plugin | Description | Category |
+|--------|-------------|----------|
+| [agentic-harness](./agentic-harness) | Stand up, assess, and maintain an agentic harness — generate project-specific agent teams and the skills they use, then assess how effectively they are used | Engineering |
+| [developer-tools](./developer-tools) | Developer environment tooling — devcontainer generation, stack detection, infrastructure config | Engineering |
+| [human-resources](./human-resources) | HR interview workflow — job descriptions, pre-screening, interview prep, evaluation, compliance | Human Resources |
+| [kaizen](./kaizen) | Continuous improvement loops — recursive optimization engine with profiles for Claude Code usage, refactoring, and process improvement | Engineering |
+| [plantuml](./plantuml) | PlantUML diagrams — policy-driven authoring, rendering, lint, validate, review, advisor, and migrate | Documentation |
+| [project-management](./project-management) | SOW writing, review, estimation, and PMI-compliant PERT analysis — integrated project management pipeline | Operations |
+| [tech-writing](./tech-writing) | Technical writing support — documentation structure, style guides, content review | Documentation |
 
 > The [`agentic-harness`](./agentic-harness) plugin is inspired by the open-source `harness` plugin by revfactory (Apache-2.0). It is an independent reimplementation under MIT, with its own structure and prose.
 
 ## Installation
 
-Add this marketplace to Claude Code:
+Add this marketplace, then install the plugins you want.
 
 ```bash
-claude plugin add MrBogomips/claude-code
+# 1. Add the marketplace
+claude plugin marketplace add MrBogomips/claude-code
+
+# 2. Install a plugin (the marketplace name is "mrbogomips")
+claude plugin install agentic-harness@mrbogomips
+```
+
+Or, inside a Claude Code session, use the slash commands:
+
+```
+/plugin marketplace add MrBogomips/claude-code
+/plugin install agentic-harness@mrbogomips
 ```
 
+Browse and manage everything interactively with `/plugin`.
+
 ## Contributing
 
 See [CONTRIBUTING.md](./CONTRIBUTING.md) for guidelines on adding or improving plugins.

agentic-harness/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "agentic-harness",
-  "version": "0.3.0",
+  "version": "0.3.1",
   "description": "Stand up, assess, and maintain an agentic harness in an existing repo. A meta-tool that generates project-specific agent teams and the skills they use, then assesses how effectively they are used. Two skills: harness-setup (build, extend, maintain) and harness-review (read-only assessment).",
   "author": {
     "name": "MrBogomips",

agentic-harness/README.md

@@ -19,16 +19,6 @@ Invoke a skill directly (`/agentic-harness:harness-setup`) or let Claude trigger
 
 `harness-setup` defaults to an **agent team** and falls back to **subagents** when the experimental team tools are unavailable. See `shared/execution-modes.md`.
 
-## Relationship to `kaizen`
-
-`kaizen` optimizes assets against measured KPIs. `agentic-harness` builds, assesses, and maintains the harness those assets live in. They compose.
-
 ## Inspired by
 
 The harness concept is inspired by prior work in the community; this is an independent implementation. Credit lives in the repository README.
-
-## Changelog
-
-- **0.3.0** — `harness-setup` adds a mandatory pre-write approval gate (Step 2b): it presents an explicit change manifest — every agent, skill, orchestrator, pointer, and tool it will create, update, remove, install, or uninstall — and writes nothing until the user formally approves the list, on every path.
-- **0.2.0** — `harness-setup` now always offers tool research (and, on an existing harness, tool maintenance) as part of the plan it presents, and accepts an optional user-provided starting context. Running still requires explicit acceptance, and per-tool adoption is unchanged. `harness-review` flags a missing tools registry as a finding.
-- **0.1.0** — Initial release: `harness-setup` and `harness-review`, shared concept docs, and the optional tool-discovery step.

agentic-harness/skills/harness-review/SKILL.md

@@ -1,7 +1,8 @@
 ---
 name: harness-review
-description: "Assess an existing agentic harness — read-only. Use to review, audit, or assess a harness; to ask how well its skills and agents are actually used; to check for drift between the files and the CLAUDE.md record; or to validate that skills trigger and agents wire up correctly. It inventories .claude/agents and .claude/skills, reads the CLAUDE.md pointer and change history, judges effective usage from project memory and the inventory, and produces a prioritized review context — what works well and what to improve — that hands off to harness-setup. This skill writes nothing; it diagnoses. To build or change a harness, use harness-setup instead."
-model: opus
+description: "Assess an existing agentic harness — read-only. Use to review, audit, or assess a harness, to judge how well its skills and agents are actually used, to check drift between the files and the CLAUDE.md record, or to validate that skills trigger and agents wire up. It produces a prioritized review context that hands off to harness-setup, and writes nothing. To build or change a harness, use harness-setup. Not for reviewing ordinary code or a pull request (that is code review)."
+model: inherit
+disallowed-tools: Write, Edit, NotebookEdit
 ---
 
 # Harness review — assess an existing harness (read-only)
@@ -27,6 +28,10 @@ to project memory. Run nothing that mutates the project. When a fix is obvious,
 the review context as a recommendation for `harness-setup`; do not apply it. The value of a
 reader is that its findings are trustworthy precisely because it changed nothing.
 
+This contract is also backed mechanically: the skill sets `disallowed-tools: Write, Edit,
+NotebookEdit`, so those tools are removed from the pool while it runs. `Bash` stays available
+for read-only inventory — the prose above still governs it; run nothing that mutates the project.
+
 ## Step 1: Inventory the harness
 
 1. List `.claude/agents/` and `.claude/skills/`; identify the orchestrator skill.

agentic-harness/skills/harness-review/references/qa-agent-guide.md

@@ -107,7 +107,7 @@ A template to fold into a QA agent for a typical web application:
 ---
 name: qa-inspector
 description: "QA verification specialist. Checks spec compliance, integration consistency, and design quality across module boundaries."
-model: opus
+model: inherit
 ---
 
 # QA inspector

agentic-harness/skills/harness-review/references/skill-testing-guide.md

@@ -80,7 +80,10 @@ A skill's description is its only trigger. Validate it with two query sets:
   unrelated query ("write a Fibonacci function") tests nothing. Build the should-NOT set from
   queries that belong to adjacent skills — especially, for this plugin's own two skills, a
   "review my harness" that must not fire `harness-setup` and an "extend my harness" that must
-  not fire `harness-review`.
+  not fire `harness-review`. Include near-misses from *external* neighbours too: "recommend
+  Claude Code automations for this repo" (belongs to `claude-code-setup`), "write a skill or
+  plugin for X" (`plugin-dev` / `skill-creator`), and "review this PR / this code" (code
+  review) — none of these should fire either harness skill.
 
 Then check for collisions: confirm the should-trigger queries don't wrongly fire a *different*
 existing skill. Where two descriptions overlap, sharpen the boundary wording in both.

agentic-harness/skills/harness-setup/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: harness-setup
-description: "Build, extend, and maintain a project's agentic harness — the agents, skills, and orchestrator under .claude/. This skill writes files. Use it to set up or build a harness, scaffold or design an agent team and the skills they use, add or change an agent or skill, update or rebuild the harness, sync it after drift, or apply a review context produced by harness-review. On explicit request it can also discover and configure external tools — MCP servers and plugins — that fit the project, and register the approved ones in the harness tools registry. Also triggers on follow-ups such as 'extend the harness', 'the harness needs a new agent', 're-run setup', 'act on the review', and 'find tools or MCPs for this project'. For read-only assessment of how well an existing harness is used, use harness-review instead — this skill is the writer, that one is the reader."
-model: opus
+description: "Build, extend, and maintain a project's agentic harness — the agents, skills, and orchestrator under .claude/. This skill writes files. Use it to set up, scaffold, extend, rebuild, or sync a harness, to add or change an agent or skill, or to apply a review context from harness-review; on request it also discovers and registers fitting MCP/plugin tools. For read-only assessment of an existing harness, use harness-review — this skill is the writer, that one is the reader. Not for authoring a single standalone skill or plugin (use plugin-dev or skill-creator), or one-shot automation recommendations (use claude-code-setup)."
+model: inherit
 ---
 
 # Harness setup — build and maintain the agent team
@@ -147,9 +147,10 @@ who it messages, who messages it, and what it claims from the shared task list.
 definition template and worked agent files are in `references/agent-design-patterns.md` and
 `references/team-examples.md`.
 
-**Model.** Set each agent's model explicitly, in both the agent file and the spawn call. A
-harness's quality tracks its agents' reasoning, so use the strongest reasoning model for
-roles that depend on judgment rather than throughput.
+**Model.** Default each agent to `model: inherit` so it follows the session's model. A
+harness's quality tracks its agents' reasoning, so for a role that depends on judgment rather
+than throughput, pin the strongest reasoning model explicitly — by its current dated id (e.g.
+`claude-opus-4-8`), not a bare `opus` alias that ages.
 
 **If the team includes a QA agent.** Use the `general-purpose` type (`Explore` is read-only
 and cannot run validation). Make its core method *cross-boundary comparison* — read both
@@ -236,7 +237,7 @@ Before calling a setup or change complete:
 - [ ] One orchestrator, with data flow, error handling, and test scenarios.
 - [ ] Execution mode is stated (team / subagent / hybrid; per-phase if hybrid), with the
       subagent fallback covered when team mode is the default.
-- [ ] Each agent's model is set explicitly.
+- [ ] Each agent's model is set deliberately (`inherit` by default; a pinned dated model id only where judgment needs it).
 - [ ] No `commands/` directory was generated.
 - [ ] No conflict with existing agents or skills.
 - [ ] Skill and orchestrator descriptions are pushy and include follow-up keywords.

agentic-harness/skills/harness-setup/references/agent-design-patterns.md

@@ -187,7 +187,7 @@ An agent leverages a skill in one of three ways:
 ---
 name: agent-name
 description: "One or two sentences on the role. List the trigger keywords."
-model: opus
+model: inherit
 ---
 
 # Agent Name — one-line role
@@ -219,5 +219,6 @@ You are an expert [role] in [domain].
 - relationships with the other agents
 ```
 
-Set `model` explicitly. For roles whose quality depends on judgment rather than throughput,
-use the strongest reasoning model available.
+Default `model` to `inherit` so the agent follows the session's model. For a role whose
+quality depends on judgment rather than throughput, pin the strongest reasoning model
+explicitly — by its current dated id (e.g. `claude-opus-4-8`), not a bare `opus` alias that ages.