getsentry / sentry-python

code-review runs a pull request through Sentry's own engineering practices — security, performance, testing, and design review together. find-bugs does the same kind of pass but scoped to the current local branch's changes rather than a PR. security-review is the OWASP-flavoured version of the same idea: injection, XSS, authentication, authorization, and cryptography issues, reported with a stated confidence level rather than a flat pass or fail.

skill-scanner is the odd one out and the most specific: it audits a different agent skill's own files before you install them, checking for prompt injection, malicious scripts, excessive permissions, secret exposure, and supply-chain risk in a SKILL.md. None of the four touches Sentry's own product surface — no sentry_sdk.init, no integrations, no event capture — this is Sentry's internal review tooling, not documentation for using Sentry.

READ THE FULL ANALYSIS＋

The README says nothing about any of this. It is the standard Sentry Python SDK readme: pip install sentry-sdk, sentry_sdk.init(...), the FastAPI, Django, Celery, and OpenAI integrations, and the 1.x-to-2.x migration guide — with no mention of skills or an agent anywhere in it. We found all four by reading the repository tree.

What it costs to start. Our dependency check splits the package 2 ready, 2 local_tool — two run on the prompt alone, and two expect a tool already present to inspect an actual git branch or repository, though the skill descriptions don't say which two.