Getting help with SafeLint

SafeLint is a community-maintained open-source project. There's no paid support tier; the channels below are best-effort, but issues are read and responded to.

Where to ask

You want to…	Go here
Report a bug (false positive, missed violation, crash, unexpected output)	Open an issue. Please include the safelint version (safelint --version), the rule code involved (e.g. SAFE101), and a minimal code snippet that reproduces it
Request a new feature (a new rule, a new AI client, a new language)	Open an issue describing the use case. For new AI clients, include the marker convention you've seen in the wild (cwd / home paths); that's what we need to add a new ClientSpec.
Ask "how do I do X with safelint"	Check README.md, CONFIGURATION.md, and AI_CLIENTS.md first; most usage questions are covered there. If still stuck, open a question issue
Contribute a fix or new feature	See CONTRIBUTING.md; three contribution paths (rule / AI client / language) are documented with checklists
Report a security vulnerability	Email the maintainer directly at srahul07@gmail.com. Please don't open a public issue for security problems; give us time to ship a fix first
Report Code of Conduct violations	Email srahul07@gmail.com. See CODE_OF_CONDUCT.md for the full process

What to include in a bug report

The faster you give us reproducible context, the faster we can help:

1. safelint version: safelint --version or pip show safelint.
2. The rule code that fired: e.g. SAFE101 (or "no rule fired" if your bug is "this should have flagged but didn't").
3. Minimal code snippet: the smallest Python file that reproduces the behaviour. Strip everything that isn't load-bearing.
4. What you expected vs. what you saw.
5. Your config: if you have a safelint.toml / pyproject.toml [tool.safelint] block, paste the relevant rule's config.
6. Your invocation: the exact safelint check ... command you ran.

A bug report with all six items typically gets answered same-day; a report missing the repro almost always gets a "could you share a minimal example?" reply first.

Documentation

• README.md, overview, install, quick start.
• CONFIGURATION.md, every CLI flag, every rule, every config option.
• AI_CLIENTS.md, the twelve supported AI clients, per-client setup, troubleshooting.
• docs/json-schema.md, the --format json output schema, for editor / plugin authors.
• CONTRIBUTING.md, how to add new rules, AI clients, or languages.

Response time expectations

This is a personal-time open-source project. Issues are typically read within a few days; substantive responses can take a week or two depending on complexity. There's no SLA. If something is genuinely urgent for you, the fastest path is usually to send a PR; that's far more likely to land quickly than a wait-for-fix issue.