- Home
- Skills
- Agents & Orchestration
- project lens
project lens
A read-only codebase explorer that explains complex projects and maps dependencies without modifying source code.
$2
Works with the AI Agents you already use
project lens
Example session with this skill installed
/project-lens how does the authentication flow work between the frontend and the auth-service?
- Read your context and instructions
- Compiled the project lens
I have analyzed the authentication flow by tracing the requests from the client-side hooks through to the backend service.
Authentication Flow Analysis
| Component | File Path | Role |
|---|---|---|
| Frontend Hook | src/hooks/useAuth.ts | Triggers login via api/auth client wrapper. |
| API Client | src/lib/api-client.ts | Attaches Bearer tokens to outbound requests. |
| Route Handler | apps/auth-service/routes/login.ts | Validates credentials against the database. |
| Token Issuer | apps/auth-service/lib/jwt.ts | Signs and issues JWTs with a 24h expiration. |
| Middleware | packages/shared/middleware/auth.ts | Verifies tokens for all protected internal routes. |
The flow starts in the frontend useAuth hook, which calls the login method in the API client. The auth-service receives the POST request, validates the user, and uses the shared JWT library to issue a token. This token is then stored in localStorage by the client and injected into the headers of subsequent requests by the api-client interceptor.
Next steps
- Review
packages/shared/middleware/auth.tsto see how roles are checked. - Verify the token expiration settings in
apps/auth-service/lib/jwt.ts. - Run a writable agent session if you need to modify the token refresh logic.
Connects securely to your tools. The creator never sees your data.
About this skill
Codebases grow fast when AI agents write the code, leaving the developer with a fragmented mental map of the system. project-lens functions as a read-only architectural guide that explains how modules connect and where features live without ever touching your source code. It eliminates the risk of accidental mutations by strictly enforcing a read-only contract, making it safe to run in parallel with active coding agents.
What it does
- Architectural mapping builds a project map to track file relationships and module boundaries.
- Contextual grounding answers questions by reading actual code and docs rather than hallucinating based on file names.
- Developer profiling adapts technical depth and language based on a persistent personal profile.
- Parallel observation reads the working tree in its current state, providing clarity during active development cycles.
- State isolation ensures no files are edited, moved, or deleted, maintaining a pure read-only environment.
How it works
- Profile Initialization conducts a one-time interview to establish your experience level and preferred technical depth.
- Project Mapping performs a read-only scan of the repository to generate a local map of the codebase.
- Information Retrieval uses
grep,find, andgit logto gather evidence for your specific questions. - Grounded Explanation delivers an answer cited with file references and architectural context.
Frameworks & tools
This tool uses standard CLI utilities including grep, find, git, and filesystem read operations. It stores metadata in Markdown files within ~/.claude/ and local .project-lens/ directories.
Why this beats prompting it yourself
Generic prompts often lead to agents making "helpful" edits or guessing at logic they haven't seen. project-lens uses a non-negotiable read-only contract and multi-layer memory to ensure answers are consistent, cited, and safe for your production code.
Use cases
- Onboarding to a new repository by generating an instant structural overview.
- Impact analysis to see which modules connect to a feature before you refactor.
- Mental map recovery when an AI agent has generated too much code to track manually.
- Safe exploration of sensitive codebases where write access must be strictly controlled.
Known limitations
This skill is strictly read-only and cannot fix bugs or implement features. It requires an initial scan to generate the project map, which may take time on very large repositories.
Read-only by design. Cannot perform code edits, builds, or migrations. Requires initial setup for profile and project map generation. Accuracy depends on existing code and documentation quality.
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 any payment.
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
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 11 days ago
Requires