agentleFS
Sign inStart free
tkollehtkolleh/skills/analyzing-architecture/SKILL.md

analyzing-architecture

Skill1 starsChanged 3 months agoMITView on GitHub

Detects repository context, maps dependencies (enterprise-specific or general), and generates/persists architectural documentation.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 6 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run gh repo view --json url or inspect the local .git/config to check the remote host.

    line 20
  2. Very specific:

    Use src search to map the ecosystem.

    line 25
  3. Distinctive:

    Contracts: Search for the organization's interface definition contracts repository for Thrift IDL, Proto schema, GraphQL, etc...

    line 26
  4. Distinctive:

    Message Queues: Search for message broker (e.g., Kafka) infrastructure-as-code repositories.

    line 27
  5. Distinctive:

    Cloud/IAM: Search for cloud infrastructure bootstrap or service provisioning repositories.

    line 28
  6. Distinctive:

    Data Pipelines: Search for ETL job or data pipeline deployer repositories.

    line 29
  7. Distinctive:

    Data Stores: Search for database schema or storage infrastructure repositories.

    line 30
  8. Very specific:

    Scan Local Entry Points: Use serena.search_code to find Main classes, server initializations, or API routes.

    line 34
  9. Very specific:

    Generate AGENT.md: Create or update the root AGENT.md.

    line 38
  10. Very specific:

    Visualize: Call the diagramming-d2 skill to generate a system map.

    line 39
  11. Very specific:

    Evidence-Based: Every dependency listed must be cited with its src search result or local file reference.

    line 43
  12. Distinctive:

    Privacy: Never save sensitive strings, keys, or secrets to memory.

    line 44

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

What I do

I analyze the current repository to determine its origin and architecture. If it is hosted on a private enterprise Sourcegraph instance, I execute a deep internal discovery protocol using the /sourcegraph-search skill. Otherwise, I perform a comprehensive local audit of language, frameworks, and public dependencies.

When to use me

Trigger this skill immediately upon entering a new repository or when the user asks "How does this system work?", "What is the architecture", "What are the system dependencies?", or "What should I know to work on this codebase?".

Operational Workflow

Phase 1: Context Detection

Run gh repo view --json url or inspect the local .git/config to check the remote host.

  • IF host is a private enterprise Sourcegraph instance: Proceed to Phase 2 (Enterprise Protocol).
  • OTHERWISE: Proceed to Phase 3 (General Protocol).

Phase 2: Enterprise Sourcegraph Discovery Protocol

Use src search to map the ecosystem. Adapt the repo patterns to your organization's Sourcegraph instance:

  1. Contracts: Search for the organization's interface definition contracts repository for Thrift IDL, Proto schema, GraphQL, etc...
  2. Message Queues: Search for message broker (e.g., Kafka) infrastructure-as-code repositories.
  3. Cloud/IAM: Search for cloud infrastructure bootstrap or service provisioning repositories.
  4. Data Pipelines: Search for ETL job or data pipeline deployer repositories.
  5. Data Stores: Search for database schema or storage infrastructure repositories.

Phase 3: General Discovery Protocol

  1. Identify Project Type: Detect primary language (e.g., package.json, build.sbt, go.mod).
  2. Scan Local Entry Points: Use serena.search_code to find Main classes, server initializations, or API routes.
  3. Map Local Dependencies: Parse dependency files to identify external services (e.g., AWS SDKs, database drivers, third-party APIs).

Phase 4: Documentation & Persistence

  1. Generate AGENT.md: Create or update the root AGENT.md.
  2. Visualize: Call the diagramming-d2 skill to generate a system map.
  3. Mandatory Memory Save: You MUST save the summarized project architecture (business goal, technical stack, and key dependencies) to your long-term memory. This ensures you do not need to re-run the full discovery in future sessions.

Constraints

  • Evidence-Based: Every dependency listed must be cited with its src search result or local file reference.
  • Privacy: Never save sensitive strings, keys, or secrets to memory.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free