agentleFS
QuixiAIQuixiAI/Hexis/skills/installed/self-inspection/SKILL.md

self-inspection

Skill623 starsChanged 10 months agoMITView on GitHub

Browse and search Hexis source code and inspect the live PostgreSQL database schema

  • Reads credentials

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 3 are distinctive, and 10 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 10

    Very specific

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

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

    Use this skill to answer questions about how Hexis is implemented, what the current checkout actually does, and what schema is running in …

    line 15
  2. Very specific:

    Use inspect_source to locate and read the relevant code.

    line 21
  3. Very specific:

    Use inspect_database_schema for current database truth.

    line 23
  4. Distinctive:

    Use the core-memory skill separately when prior decisions or experiences are relevant.

    line 25
  5. Distinctive:

    Follow references across modules when needed.

    line 40
  6. Very specific:

    Start with overview only when the relevant object is unknown.

    line 45
  7. Very specific:

    Use search for a table, column, view, or stored-function name fragment.

    line 46
  8. Very specific:

    Use describe_relation for columns, defaults, constraints, indexes, and view definitions.

    line 47
  9. Very specific:

    Use get_function to inspect the actual stored definition and overloads.

    line 49
  10. Very specific:

    Use review_recent_changes to see what changed about your own substrate — migrations, code rebuilds, prompt edits, and operator config …

    line 58
  11. Very specific:

    When tool calls, heartbeat steps, integrations, or memory retrieval have been failing, use self_repair to list unresolved defect reports, …

    line 60
  12. Distinctive:

    Do not claim access to model weights, provider infrastructure, hardware, or secrets unless another explicit capability supplies that …

    line 65
  13. Very specific:

    Do not expose credentials, .env contents, or unrelated private files.

    line 67

≈ “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

Self Inspection

Use this skill to answer questions about how Hexis is implemented, what the current checkout actually does, and what schema is running in PostgreSQL.

Evidence Order

  1. Inspect the source or live schema before making implementation claims.
  2. Use inspect_source to locate and read the relevant code. Start with search, then read only the necessary line ranges.
  3. Use inspect_database_schema for current database truth. Prefer describe_relation or get_function after a narrow search.
  4. Use the core-memory skill separately when prior decisions or experiences are relevant. Memory is historical evidence; source and live schema are implementation evidence.
  5. Reconcile differences explicitly. Baseline SQL files describe a fresh database, migrations evolve existing databases, and the live schema is the authority for what is running now.

Source Method

  • inspect_source(action="list", path="...", file_pattern="...") discovers repository files.
  • inspect_source(action="search", query="...", path="...", file_pattern="...") finds definitions and call sites.
  • inspect_source(action="read", path="...", offset=..., limit=...) reads a bounded, line-numbered range.
  • Follow references across modules when needed. Do not infer behavior from a filename or one isolated function.

Schema Method

  • Start with overview only when the relevant object is unknown.
  • Use search for a table, column, view, or stored-function name fragment.
  • Use describe_relation for columns, defaults, constraints, indexes, and view definitions.
  • Use get_function to inspect the actual stored definition and overloads.
  • The schema tool is metadata-only and does not accept arbitrary SQL.

Reporting

  • Distinguish facts observed in source, facts observed in the live schema, and inferences.
  • Cite repository-relative paths and function/relation names.
  • State when the running schema differs from baseline files or migrations.
  • Use review_recent_changes to see what changed about your own substrate — migrations, code rebuilds, prompt edits, and operator config decisions.
  • When tool calls, heartbeat steps, integrations, or memory retrieval have been failing, use self_repair to list unresolved defect reports, diagnose one with the source/schema evidence gathered here, and mark it resolved only after verification. It records and reasons about defects; it never silently edits source code.
  • Do not claim access to model weights, provider infrastructure, hardware, or secrets unless another explicit capability supplies that evidence.
  • Do not expose credentials, .env contents, or unrelated private files.

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.