docs: rename 0.5.0 task to "agent skill" — clarify multi-agent support

`npx skills add` (vercel-labs/skills) targets 45+ agents, not just Claude
Code. Reword roadmap entry, README blurb, AGENTS tree comment, and the
SessionStart banner accordingly. The optional auto-check hook is the only
piece that's still host-specific; task file now lists Claude Code and
Codex CLI as the two agents with a SessionStart hook today.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: hsnice16

.claude/settings.json

@@ -6,7 +6,7 @@
         "hooks": [
           {
             "type": "command",
-            "command": "printf '\\n📦 Agent Friendly Code — current release: 0.4.0\\n  • Read AGENTS.md for conventions, CONTRIBUTING.md for the PR workflow.\\n  • Roadmap: 0.5.0 (quick wins — history-aware signals + PR score-diff + Claude Code skill) → 0.6.0 (auto-refresh + smarter matching — webhook rescoring + alternatives v2) → 0.7.0 (maintainer ownership + at-scale discovery — OAuth opt-out + package overlay at scale) → 1.0.0 (production cut — Postgres + at-scale indexing + benchmark harness).\\n  • Changelog rule: user-facing capabilities only. Codebase hygiene (CI / linter / tests / CONTRIBUTING) does NOT go in lib/changelog.ts.\\n'"
+            "command": "printf '\\n📦 Agent Friendly Code — current release: 0.4.0\\n  • Read AGENTS.md for conventions, CONTRIBUTING.md for the PR workflow.\\n  • Roadmap: 0.5.0 (quick wins — history-aware signals + PR score-diff + agent skill) → 0.6.0 (auto-refresh + smarter matching — webhook rescoring + alternatives v2) → 0.7.0 (maintainer ownership + at-scale discovery — OAuth opt-out + package overlay at scale) → 1.0.0 (production cut — Postgres + at-scale indexing + benchmark harness).\\n  • Changelog rule: user-facing capabilities only. Codebase hygiene (CI / linter / tests / CONTRIBUTING) does NOT go in lib/changelog.ts.\\n'"
           }
         ]
       }

AGENTS.md

@@ -97,7 +97,7 @@ tasks/
   0.2.0/                  # released — dogfood complete (tests, self-score, row-click)
   0.3.0/                  # released — embeddable scores + broader coverage (badge, more agents, alternatives, package lookup)
   0.4.0/                  # released — credible scores + discoverability (docs-cited rationales + agent-specific signals + About/llms.txt/OG)
-  0.5.0/                  # planned — quick wins (history-aware signals + PR score-diff action + Claude Code skill)
+  0.5.0/                  # planned — quick wins (history-aware signals + PR score-diff action + agent skill)
   0.6.0/                  # planned — auto-refresh + smarter matching (webhook rescoring + alternatives v2)
   0.7.0/                  # planned — maintainer ownership + at-scale discovery (OAuth opt-out + package overlay at scale)
   1.0.0/                  # planned — production cut (Postgres + at-scale indexing + benchmark harness)

README.md

@@ -172,7 +172,7 @@ See `/roadmap` in the running app or the per-version `tasks/` folders for the fu
 
 Versions are sequenced cheap-first so the highest-impact small additions don't get gated on heavy infra:
 
-- **0.5.0 — quick wins**: history-aware signals (maintenance recency, commit velocity, contributor activity) + a GitHub Action that comments the score delta on every PR + a Claude Code skill (with public `/api/score` lookup) that recommends a model for the active repo. No new infra.
+- **0.5.0 — quick wins**: history-aware signals (maintenance recency, commit velocity, contributor activity) + a GitHub Action that comments the score delta on every PR + a portable agent skill (Claude Code / Codex / Cursor / Cline / Copilot / 40+ others, with public `/api/score` lookup) that recommends a model for the active repo. No new infra.
 - **0.6.0 — auto-refresh + smarter matching**: webhook-driven rescoring (keep scores fresh on every push) + alternatives via README embeddings (cross-language matches the v0.3.0 SQL heuristic misses).
 - **0.7.0 — maintainer ownership + at-scale discovery**: OAuth opt-out / claim flow for maintainers + at-scale package overlay (per-registry leaderboards + userscript that renders the badge inline on npmjs.com / PyPI / crates.io).
 - **1.0.0 — production cut**: Postgres migration for concurrent writers + auto-discovered crawl (target 10k repos) + benchmark harness that derives per-model weights from measured agent success. From here on, breaking API changes require a MAJOR bump.

lib/roadmap.ts

@@ -29,10 +29,10 @@ export const ROADMAP: RoadmapVersion[] = [
         summary: "GitHub Action that comments score delta on every PR.",
       },
       {
-        title: "Claude Code skill",
-        taskFile: "tasks/0.5.0/03-claude-code-skill.md",
+        title: "Agent skill",
+        taskFile: "tasks/0.5.0/03-agent-skill.md",
         summary:
-          "Skill (installable via npx skills add) that checks the active repo's score and recommends a Claude model. Ships with the /api/score lookup endpoint and a UI integration page with a SessionStart hook snippet.",
+          "Portable skill (installable via npx skills add into Claude Code, Codex, Cursor, Cline, Copilot, …) that checks the active repo's score and recommends a model. Ships with the /api/score lookup endpoint and a UI integration page with auto-check hook snippets for the agents that support them.",
       },
     ],
   },

tasks/0.5.0/03-agent-skill.md

@@ -0,0 +1,68 @@
+# 03 · Agent skill (with public score lookup)
+
+**Status**: planned
+
+## Goal
+
+Ship an agent skill that fetches the active repo's score and recommends a model, plus the public lookup endpoint it depends on and a UI page that explains the integration. Bundled because the endpoint has no other consumer in 0.5.0 and the three pieces only make sense together.
+
+The skill itself is portable — installable via `npx skills add hsnice16/agent-friendly-code`. The [vercel-labs/skills](https://github.com/vercel-labs/skills) CLI supports Claude Code, Codex, Cursor, Cline, Copilot, and 40+ other agents, and writes to the right per-agent directory based on the `--agent` flag.
+
+## Pieces
+
+### 1. Public lookup endpoint
+
+`GET /api/score?host=github&repo=<owner>/<name>`
+
+- 200 with `{ repo, signals, modelScores }` — same shape as `/api/repo/[id]` so consumers share a type.
+- 404 with `{ error: "not_indexed" }` when not found — explicit so the skill can branch on it cleanly.
+- 400 on malformed `repo`. `host` defaults to `github`.
+- Thin wrapper around `getRepoByHostOwnerName` in `lib/db.ts`. No on-demand scoring; no GitHub token usage at request time.
+
+### 2. The skill
+
+```text
+skills/
+  agent-friendly/
+    SKILL.md
+```
+
+Frontmatter (per vercel-labs/skills convention — agent-neutral):
+
+```yaml
+---
+name: agent-friendly
+description: Check the active repo's agent-friendliness score and recommend a model. Use when the user asks "is this repo a mess?", "which model should I use here?", or invokes /agent-friendly.
+---
+```
+
+Body: instructions for the agent to (a) resolve `owner/name` from the git remote, (b) `GET /api/score` against the deployed instance, (c) on 200, render score + recommended model, (d) on 404, say so plainly and skip the recommendation, (e) on first invocation, surface the auto-check hook snippet so the user can opt into automatic checks (when their agent supports session-start hooks). Score → model mapping lives in this file (high score → frontier model, low score → small/fast model); exact thresholds and model IDs land during implementation. Recommendation is provider-neutral — the same score buckets map to Opus/Sonnet/Haiku, GPT-5/4o-mini, etc.
+
+### 3. UI page
+
+New route (e.g. `/skill` or `/integrations/agents`):
+
+- One-paragraph pitch.
+- Install command: `npx skills add hsnice16/agent-friendly-code` (link to the [vercel-labs/skills](https://github.com/vercel-labs/skills) agent list so users know it isn't Claude Code-only).
+- Manual invocation: `/agent-friendly` (or whatever invocation the host agent uses).
+- Optional auto-check hook snippets — shown for the agents that support session-start hooks today:
+  - **Claude Code**: `SessionStart` hook in `.claude/settings.json`
+  - **Codex CLI**: `SessionStart` matcher in `.codex/hooks.json` ([docs](https://developers.openai.com/codex/hooks))
+  - **Cursor / Cline / others**: no session-start hook event yet — the skill itself fires when the user mentions the repo, or they can paste the snippet into `.cursorrules` / `.clinerules` as a static instruction.
+- Linked from the homepage and added to `lib/roadmap.ts`.
+
+The hook snippets on the UI page and the skill's first-invocation nudge must stay in sync — same content, multiple surfaces.
+
+## Acceptance
+
+- `GET /api/score?host=github&repo=owner/name` returns the documented shape; 404 body distinguishes `not_indexed` from generic errors.
+- `npx skills add hsnice16/agent-friendly-code` installs into the right per-agent directory (`.claude/skills/agent-friendly/`, `.codex/skills/agent-friendly/`, `.cursor/skills/agent-friendly/`, …).
+- `/agent-friendly` in a session inside a git repo prints the score + model recommendation, or a clean "not indexed" message.
+- UI page is reachable from the homepage; install command + hook snippets match the skill's output, and the page makes clear which agents support the auto-check hook today.
+- README gains a short "Public API" section pointing skill / hook authors at `/api/score`.
+
+## Out of scope
+
+- On-demand scoring of unindexed repos.
+- Programmatic model switching. The skill recommends; the user runs `/model` (or their agent's equivalent) themselves.
+- Per-agent forks of the skill body. One `SKILL.md` for all agents; only the optional hook snippet differs by host.

tasks/0.5.0/03-claude-code-skill.md

@@ -2,10 +2,10 @@
 
 **Status**: planned
 
-The cheapest items on the roadmap, paired so they can ship in one cut. History-aware signals extend the existing host-API clients with ~3 new signal files; the PR-diff GitHub Action is a thin wrapper that calls the badge / API endpoints we already ship; the Claude Code skill bundles a small lookup endpoint, a `SKILL.md` shipped from this repo, and a UI integration page. No new infra, no new deps of consequence — high-impact additions that don't need a heavy release.
+The cheapest items on the roadmap, paired so they can ship in one cut. History-aware signals extend the existing host-API clients with ~3 new signal files; the PR-diff GitHub Action is a thin wrapper that calls the badge / API endpoints we already ship; the agent skill bundles a small lookup endpoint, a portable `SKILL.md` shipped from this repo, and a UI integration page. No new infra, no new deps of consequence — high-impact additions that don't need a heavy release.
 
 ## Tasks
 
 - [01-history-aware-signals.md](./01-history-aware-signals.md) — extend the scorer with maintenance recency, commit velocity, and contributor activity. Hybrid fetch (shallow clone for files, host API for history) — degrades gracefully without a token.
 - [02-score-diff-on-pr.md](./02-score-diff-on-pr.md) — GitHub Action that comments the score delta on every PR using the existing `/api/repo/:id` and `/badge/...svg` endpoints.
-- [03-claude-code-skill.md](./03-claude-code-skill.md) — Claude Code skill (installable via `npx skills add`) that fetches the active repo's score and recommends a model, bundled with the public `/api/score` lookup endpoint it depends on and a UI integration page with the SessionStart hook snippet.
+- [03-agent-skill.md](./03-agent-skill.md) — portable agent skill (installable via `npx skills add` into Claude Code, Codex, Cursor, Cline, Copilot, …) that fetches the active repo's score and recommends a model, bundled with the public `/api/score` lookup endpoint it depends on and a UI integration page with auto-check hook snippets for the agents that support them.