feat(hooks): opt-in skill router with evaluation (split from #2788)

.claude-plugin/marketplace.json

@@ -11,7 +11,7 @@
     {
       "name": "ecc",
       "source": "./",
-      "description": "Harness-native ECC operator layer - 68 agents, 286 skills, 94 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses",
+      "description": "Harness-native ECC operator layer - 68 agents, 287 skills, 95 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses",
       "version": "2.2.1",
       "author": {
         "name": "Affaan Mustafa",

.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "ecc",
   "version": "2.2.1",
-  "description": "Harness-native ECC plugin for engineering teams - 68 agents, 286 skills, 94 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses",
+  "description": "Harness-native ECC plugin for engineering teams - 68 agents, 287 skills, 95 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses",
   "author": {
     "name": "Affaan Mustafa",
     "url": "https://x.com/affaanmustafa"

AGENTS.md

@@ -1,6 +1,6 @@
 # Everything Claude Code (ECC) — Agent Instructions
 
-This is a **production-ready AI coding plugin** providing 68 specialized agents, 286 skills, 94 commands, and automated hook workflows for software development.
+This is a **production-ready AI coding plugin** providing 68 specialized agents, 287 skills, 95 commands, and automated hook workflows for software development.
 
 **Version:** 2.2.1
 
@@ -154,8 +154,8 @@ Troubleshoot failures: check test isolation → verify mocks → fix implementat
 
 ```
 agents/          — 68 specialized subagents
-skills/          — 286 workflow skills and domain knowledge
-commands/        — 94 slash commands
+skills/          — 287 workflow skills and domain knowledge
+commands/        — 95 slash commands
 hooks/           — Trigger-based automations
 rules/           — Always-follow guidelines (common + per-language)
 scripts/         — Cross-platform Node.js utilities

COMMANDS-QUICK-REF.md

@@ -219,6 +219,7 @@ These lifecycle commands are also available through the `ecc` CLI.
 | `/projects` | List known projects and their instinct statistics |
 | `/project-init` | Detect a project's stack and produce a dry-run ECC onboarding plan |
 | `/harness-audit` | Audit the agent harness configuration for reliability and cost |
+| `/plugin-profiles` | Generate a slim ECC plugin carrier so a project loads only the surface it needs |
 | `/model-route` | Route a task to the right model (Haiku / Sonnet / Opus) |
 | `/pm2` | PM2 process manager initialisation |
 | `/setup-pm` | Configure package manager (npm / pnpm / yarn / bun) |

README.md

@@ -162,13 +162,13 @@ Instead of rebuilding that process in every prompt, you install it once and make
 
 ECC is MIT-licensed open source. It works best with Claude Code today, has a supported Codex sync path, and provides capability-limited adapters for Cursor, OpenCode, Gemini, Zed, GitHub Copilot, Antigravity, Qwen, and other harnesses. See the [support status matrix](#platform-support) before assuming feature parity.
 
-Access to 68 agents, 286 skills, and 94 legacy command shims, plus hooks, rules, memory, continuous learning, and AgentShield security scanning. The agents are specialized for planning, review, build repair, security, architecture, and domain work.
+Access to 68 agents, 287 skills, and 95 legacy command shims, plus hooks, rules, memory, continuous learning, and AgentShield security scanning. The agents are specialized for planning, review, build repair, security, architecture, and domain work.
 
 | Included         |       Count | What it gives you                                                                    |
 | ---------------- | ----------: | ------------------------------------------------------------------------------------ |
 | Agents           |   68 agents | Planning, review, build repair, security, architecture, and domain work              |
-| Skills           |  286 skills | TDD, research, security, docs, frontend, data, ML, operations, and more              |
-| Commands         | 94 commands | Convenient entry points while ECC moves to a skills-first surface                    |
+| Skills           |  287 skills | TDD, research, security, docs, frontend, data, ML, operations, and more              |
+| Commands         | 95 commands | Convenient entry points while ECC moves to a skills-first surface                    |
 | Hooks and memory |     Runtime | Enforcement, session summaries, continuous learning, instincts, and context controls |
 | Rules            |   Selective | Always-loaded standards you choose by language or project                            |
 | AgentShield      |    Included | Scanning for prompts, hooks, MCP config, permissions, secrets, and agent files       |

README.zh-CN.md

@@ -196,7 +196,7 @@ Copy-Item -Recurse rules/typescript "$HOME/.claude/rules/"
 /plugin list ecc@ecc
 ```
 
-**完成！** 你现在可以使用 68 个代理、286 个技能和 94 个命令。
+**完成！** 你现在可以使用 68 个代理、287 个技能和 95 个命令。
 
 ### multi-* 命令需要额外配置
 

agent.yaml

@@ -110,6 +110,7 @@ skills:
   - perl-testing
   - plankton-code-quality
   - plan-canvas
+  - plugin-profiles
   - plan-orchestrate
   - postgres-patterns
   - product-lens
@@ -220,6 +221,7 @@ commands:
   - plan
   - plan-canvas
   - plan-prd
+  - plugin-profiles
   - pm2
   - projects
   - promote

commands/plugin-profiles.md

@@ -0,0 +1,55 @@
+---
+description: Generate and manage slim ECC profile plugin carriers - list profiles, plan the listing ledger and capability decision, generate a receipted plugin, and activate it per project.
+argument-hint: "[list | plan <profile> | generate <profile> | activate <plugin-name>]"
+---
+
+# Plugin Profiles Command
+
+Manage slim ECC profile plugin carriers from inside Claude Code. A carrier is
+a standalone plugin generated from an install selection: it lists only the
+selected skills, agents, and commands, keeps the rest of the skill catalog
+reachable on demand inside the plugin, and records how it was built in
+`ecc-profile.json`.
+
+This command is a thin entry point over the `plugin-profiles` skill. Follow
+that skill for the full workflow and rules.
+
+## What This Command Does
+
+1. `list` - show the available install-profile projections and their module counts.
+2. `plan <profile>` - report the context surface, the token ledger with its
+   method label and budget verdict, and the capability decision.
+3. `generate <profile>` - dry run first, resolve every blocker with the user,
+   then generate and show the receipt digests.
+4. `activate <plugin-name>` - offer to write the per-project opt-in to
+   `.claude/settings.json`, merging rather than overwriting, and only after
+   showing the JSON and getting confirmation.
+
+Two things the skill is strict about, repeated here because they are the easy
+mistakes: a narrow context selection never authorizes the hook runtime, so
+never choose `--hooks` for the user; and the default ledger over-counts, so
+"OVER budget" may be a false positive that `--measure provider` clears.
+
+Profile ids (`minimal`, `developer`, `opencode`, ...) are **install-profile
+projections**, not context profiles: ECC has no canonical context-profile
+registry yet, so every receipt records `registry: install-profiles@unbound`.
+
+## Example
+
+```
+User: /plugin-profiles generate opencode
+
+Assistant: (runs plan, then generate --dry-run)
+  Ledger 8,323 tokens - OVER the 8,000 budget. That estimate over-counts by
+  design; --measure provider would give the real number. No hook runtime in
+  this selection, so no capability decision is needed.
+  Proceed with --allow-over-budget, raise the budget, or narrow the selection?
+User: allow it
+Assistant: (generates, reports path + receipt digests, offers activation)
+```
+
+## Related
+
+- `plugin-profiles` skill - full workflow, blocker table, receipt reading, rules
+- `docs/PLUGIN-PROFILES.md` - design rules, fail-closed behaviour, receipt schema
+- Source: `scripts/plugin-profiles.js`, `scripts/lib/plugin-profiles/`

docs/COMMAND-AGENT-MAP.md

@@ -6,6 +6,7 @@ This document lists each slash command and the primary agent(s) or skills it inv
 |---------|------------------|--------|
 | `/plan` | planner | Implementation planning before code |
 | `/plan-canvas` | — (skill: plan-canvas) | Browser review canvas for plan artifacts: annotate, chat, approve/request changes |
+| `/plugin-profiles` | — (skill: plugin-profiles) | Generate, inspect, and activate slim ECC plugin carriers |
 | `/tdd` | tdd-guide | Test-driven development |
 | `/code-review` | code-reviewer | Quality and security review |
 | `/build-fix` | build-error-resolver | Fix build/type errors |

docs/COMMAND-REGISTRY.json

@@ -1,6 +1,6 @@
 {
   "schemaVersion": 1,
-  "totalCommands": 94,
+  "totalCommands": 95,
   "commands": [
     {
       "command": "aside",
@@ -661,6 +661,17 @@
       ],
       "path": "commands/plan.md"
     },
+    {
+      "command": "plugin-profiles",
+      "description": "Generate and manage slim ECC profile plugin carriers - list profiles, plan the listing ledger and capability decision, generate a receipted plugin, and activate it per project.",
+      "type": "planning",
+      "primaryAgents": [],
+      "allAgents": [],
+      "skills": [
+        "plugin-profiles"
+      ],
+      "path": "commands/plugin-profiles.md"
+    },
     {
       "command": "pm2",
       "description": "Analyze a project and generate PM2 service commands for detected frontend, backend, or database services.",
@@ -1031,7 +1042,7 @@
       "build": 2,
       "general": 10,
       "orchestration": 11,
-      "planning": 2,
+      "planning": 3,
       "refactoring": 1,
       "review": 15,
       "testing": 53