/codebase-search
Preferred local codebase-understanding workflow for Pi and Codex. Start with codebase_context before shell search or broad reads, then use specialized semantic and graph tools.
$ npx -y skills add Helweg/open-codebase-index --skill codebase-search --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/codebase-search
Context preview
The summary Claude sees to decide when to auto-load this skill.
Preferred local codebase-understanding workflow for Pi and Codex. Start with codebase_context before shell search or broad reads, then use specialized semantic and graph tools.
SKILL.md
codebase-search.SKILL.mdname: codebase-search
description: Preferred local codebase-understanding workflow for Pi and Codex. Start with codebase_context before shell search or broad reads, then use specialized semantic and graph tools.
Codebase Search Skill
Use this skill when you need local repository knowledge before web lookup.
Core workflow
1. Run `index_status` when index readiness or freshness is unknown. 2. Use `codebase_context(query, ...)` before shell search, grep, or broad file reads. Pass `symbol` for an authoritative definition or `from` + `to` for a dependency path. 3. Use `codebase_peek(query, ...)` for specialized metadata-only conceptual lookup. 4. Use `codebase_search(query, ...)` when you need full code context. 5. Use `implementation_lookup(query)` for known-symbol definitions and `call_graph` / `call_graph_path` for execution flow. 6. Use `find_similar(code)` for duplicate patterns and refactor planning.
If results are weak, run `index_status` (check readiness) and `index_codebase`.
Tool Priority
- `codebase_context` as the preferred first repository tool and unified router.
- `codebase_peek` for specialized discovery (fastest, cheap tokens).
- `codebase_search` for exact implementation review.
- `find_similar` for pattern matching and duplication.
- `call_graph` and `call_graph_path` for execution flow.
- `index_codebase` (force/estimate/verbose) for first-time or stale indexes.
- `index_status`, `index_health_check`, `index_metrics`, `index_logs` for operational checks.
Suggested Commands
1. `codebase_peek("payment processing flow")` 2. `codebase_search("payment processing flow")` 3. `call_graph("chargeCard", "callees")` 4. `find_similar("function validate(data)")` 5. `implementation_lookup("validate")`
Additional Notes
- Use `grep` for exact identifiers and tiny, deterministic lookups.
- Use `websearch` only when local tools return no results and docs are likely missing.
- Prefer `codebase_peek` before `codebase_search` to avoid high token usage.
Read more
name: codebase-search description: Preferred local codebase-understanding workflow for Pi and Codex. Start with codebase_context before shell search or broad reads, then use specialized semantic and graph tools.
Codebase Search Skill
Use this skill when you need local repository knowledge before web lookup.
Core workflow
1. Run `index_status` when index readiness or freshness is unknown. 2. Use `codebase_context(query, ...)` before shell search, grep, or broad file reads. Pass `symbol` for an authoritative definition or `from` + `to` for a dependency path. 3. Use `codebase_peek(query, ...)` for specialized metadata-only conceptual lookup. 4. Use `codebase_search(query, ...)` when you need full code context. 5. Use `implementation_lookup(query)` for known-symbol definitions and `call_graph` / `call_graph_path` for execution flow. 6. Use `find_similar(code)` for duplicate patterns and refactor planning.
If results are weak, run `index_status` (check readiness) and `index_codebase`.
Tool Priority
- `codebase_context` as the preferred first repository tool and unified router.
- `codebase_peek` for specialized discovery (fastest, cheap tokens).
- `codebase_search` for exact implementation review.
- `find_similar` for pattern matching and duplication.
- `call_graph` and `call_graph_path` for execution flow.
- `index_codebase` (force/estimate/verbose) for first-time or stale indexes.
- `index_status`, `index_health_check`, `index_metrics`, `index_logs` for operational checks.
Suggested Commands
1. `codebase_peek("payment processing flow")` 2. `codebase_search("payment processing flow")` 3. `call_graph("chargeCard", "callees")` 4. `find_similar("function validate(data)")` 5. `implementation_lookup("validate")`
Additional Notes
- Use `grep` for exact identifiers and tiny, deterministic lookups.
- Use `websearch` only when local tools return no results and docs are likely missing.
- Prefer `codebase_peek` before `codebase_search` to avoid high token usage.
Search a codebase by meaning, then follow the result into definitions, callers, and dependency paths. open-codebase-index is a local semantic code index for OpenCode, Jcode, Pi, Codex, Claude Code, and other MCP clients.
Repo: Helweg/open-codebase-index