Works with the AI Agents you already use
codenav
Navigate and analyze codebases as a graph to find call sites, dependencies, and change impact.
Free
codenav
Example session with this skill installed
Analyze the impact of changing the storage.Save method in this Rust repo and tell me which tests to run.
- Read your context and instructions
- Compiled the codenav
I have analyzed the call graph for storage.Save. The change has a moderate impact radius, primarily affecting the API handler and the background worker modules.
Impact Analysis: storage.Save
| Symbol | File | Relation | Confidence |
|---|---|---|---|
storage.Save | src/storage/mod.rs | Target | Exact |
api.v1.UploadHandler | src/api/v1/upload.rs | Caller | Exact |
worker.CleanupTask | src/worker/cleanup.rs | Caller | Inferred |
models.UserMetadata | src/models/user.rs | Dependency | Exact |
Impact Summary
Changing the signature of storage.Save will require updates in src/api/v1/upload.rs where the multipart form data is processed. The worker.CleanupTask also calls this method during its nightly reconciliation loop. No circular dependencies were detected during this trace.
Recommended Tests
Based on the graph traversal, the following test suites cover the affected paths:
tests/api_upload_integration.rssrc/storage/tests.rssrc/worker/cleanup_test.rs
Next steps
- Run
cargo testspecifically for the identified files to establish a baseline. - Use
read_codeonsrc/api/v1/upload.rsto see how theSaveresult is currently handled. - Apply the signature change and use
diff_impactto verify no other symbols were affected.
Connects securely to your tools. The creator never sees your data.
About this skill
Grepping for symbols and tracing dependencies manually is slow and error-prone in large codebases. This skill uses the codenav MCP server to index your repository as a graph, letting you query code relationships with high precision instead of fuzzy text searches.
What it does
- Graph-based navigation: Maps files, symbols, calls, imports, and inheritance across the entire project.
- Impact analysis: Identifies what code will break if a specific symbol or function is modified.
- Path discovery: Finds the direct or indirect connections between two arbitrary points in the code.
- Dependency auditing: Detects circular dependencies at the module level and identifies tightly coupled components.
- Automated test selection: Determines exactly which tests to run based on recent file changes or staged diffs.
How it works
- Installation: Install the
codenavbinary via Cargo or GitHub releases to handle local indexing. - Indexing: Run the
servecommand to build a local code graph in the.codenav/directory. - Exploration: Use the
exploretool to ask structural questions like "how are users created" to get a mapped overview. - Deep dive: Chain specific tools like
get_neighborsorfind_pathsto trace call hierarchies and dependencies.
Frameworks & tools
This skill supports Rust, Python, TypeScript, JavaScript, Go, and Swift repositories. It integrates with MCP-compatible clients including Claude Code, Cursor, and Claude Desktop.
Why this beats prompting it yourself
Standard LLM prompts rely on the limited file context you provide manually. This skill gives the agent access to a pre-indexed graph of the whole repository, allowing it to "see" call sites and inheritance chains that are not currently open in your editor.
Use cases
- Onboarding: Generate a module-level overview to understand a new repository's architecture.
- Refactoring: Run an impact analysis before changing a core API to find all downstream callers.
- Debugging: Trace paths between a data source and a sink to find where state is being mutated.
- CI optimization: Identify the minimal set of tests required to validate a specific pull request.
Known limitations
Does not parse string literals, comments, or config files. Dynamic dispatch and reflection may result in missing edges. Requires local binary installation.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Requires local codenav binary. Limited to Rust, Python, TS/JS, Go, and Swift. Cannot resolve dynamic dispatch or reflection with 100% certainty.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 10 days ago
- Free with an Agensi account
- Software
Python