sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Debug a failure - test failures, runtime errors, unexpected behaviour, or build breaks. Use when the user reports a bug, a test fails, an error trace appears, output is wrong, or behaviour is unexplained.
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Select and optimize embedding models for semantic search and RAG applications. Use when choosing embedding models, implementing chunking strategies, or optimizing embedding quality for specific domains. Branches: compare-models (evaluate candidates), optimize-chunking (tune chunk strategy), fine-tune-embeddings…
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Improve AI applications with evaluation-driven development. Define eval criteria, instrument the app, build golden datasets, run pixie tests, and produce actionable improvements. Use when the user asks to set up QA, add tests/evaluations, benchmark, or improve quality for any Python project that calls an LLM.
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Pythonic idioms, PEP 8 standards, type hints, and best practices for Python code. Use when writing Python code, reviewing type annotations, choosing concurrency patterns, structuring imports, designing packages, or optimizing performance.
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Python testing strategies using pytest, TDD methodology, fixtures, mocking, parametrization, and coverage requirements. Use when the user asks to "write tests", "set up pytest", "mock dependencies", "test async code", or wants TDD guidance.
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Create, rewrite, or audit a project README. Use when the user asks for a README, documentation overhaul, project description, or getting-started guide. Also use when another skill ships a feature that needs documenting for external readers.
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Scan codebases for security vulnerabilities - injection flaws, auth bypass, secrets exposure, weak crypto, insecure deps, and business logic bugs. Use when asked to review code for security issues, audit for vulnerabilities, check for SQLi/XSS/command injection, find exposed API keys or hardcoded secrets, or any…
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Audit skill compliance across 6 dimensions. Use when the user asks to check compliance, audit skills, run compliance checks, or verify skill quality. Branches: audit-all (run all checks in order), audit (run specific check by name), fix (auto-fix a failing check).
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Reference for writing and editing skills well — the vocabulary and principles that make a skill predictable.
sagar-shirwalkar/collibra-atlas
Skill Claude CodeCodex
Search the web and extract information from pages. Use when the user asks a question that requires current information, wants research on a topic, needs documentation lookup, or when you lack the knowledge or recency to answer from training data.
sagar-shirwalkar/collibra-atlas
Instructions file CodexOpenCode
AGENTS.md instructions for sagar-shirwalkar/collibra-atlas, covering collibra atlas — agent guide, project in brief, skill catalog, convention: disable-model-invocation: true and scanning for new skills.
sagar-shirwalkar/collibra-atlas
MCP server Claude CodeCodexCursor +2
MCP server "collibra-atlas" as configured in sagar-shirwalkar/collibra-atlas. Runs locally from the collibra-atlas Python package.