feat(skill-detect): intent-to-skill injection + skill install CLI (#231, #233)

[getappz]

Summary

• feat(github): paginate flare_git list actions instead of truncating at 30 #233 — agentflare skill install/search/list/remove/registry CLI, backed by the skill crate (git clone + SKILL.md discovery). Home resolution now goes through crate::paths::home() for consistency/testability instead of raw dirs::home_dir().
• refactor: consolidate agent-config path + JSON read/write helpers #231 — intent classifier (skill_detect::classify) + skill_detect MCP tool + session-start and per-turn hook injection that map a prompt to a TaskType and rank matching skills from skill_registry. Ranking blends BM25 with cosine similarity from the existing MiniLM ONNX pipeline (memory::engine::embed_query/embed_doc, same 0.4/0.6 weighting as memory::search::search_hybrid); degrades to BM25-only when the semantic feature/model isn't available.

Test plan

• cargo test --bin agentflare skill — 27 tests, incl. new rerank_with_embeddings unit tests and an end-to-end local-source skill install test that actually copies a fixture SKILL.md into the install dir
• cargo check --features semantic — semantic path compiles
• cargo clippy --bin agentflare -- -D warnings — clean on every file this PR touches
• rustfmt --edition 2024 --check — clean on every file this PR touches

[coderabbitai]

Warning

Review limit reached

@getappz, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 27 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 6582524c-541c-4a01-9e8c-3c44fc6ecee0

📥 Commits

Reviewing files that changed from the base of the PR and between 63a22d6 and 6d810b4.

⛔ Files ignored due to path filters (1)

• Cargo.lock is excluded by !**/*.lock

📒 Files selected for processing (10)

• Cargo.toml
• crates/skill-registry/Cargo.toml
• crates/skill-registry/src/frontmatter.rs
• src/cli/mod.rs
• src/cli/skill.rs
• src/hook.rs
• src/main.rs
• src/mcp_server.rs
• src/mcp_server/types.rs
• src/skill_detect.rs

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch item-231-233-skill-detect

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}