Skip to content

docs: write README in maintainer voice #4

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Labeeb2339 merged 2 commits into from
Jul 22, 2026
Merged

docs: write README in maintainer voice #4

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
eefe1da
docs: write README in maintainer voice
Labeeb2339 Jul 22, 2026
f9e34fe
ci: run required audits on every pull request
Labeeb2339 Jul 22, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 0 additions & 10 deletions .github/workflows/dependency-audit.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,17 +3,7 @@ name: Dependency audit
on:
push:
branches: [main]
paths:
- pyproject.toml
- web/package.json
- web/package-lock.json
- .github/workflows/dependency-audit.yml
pull_request:
paths:
- pyproject.toml
- web/package.json
- web/package-lock.json
- .github/workflows/dependency-audit.yml
schedule:
- cron: "19 3 * * 1"
workflow_dispatch:
Expand Down
10 changes: 5 additions & 5 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,11 +8,11 @@

![SFTGuard confirmatory dashboard](docs/assets/sftguard-confirmatory-dashboard.png)

SFTGuard catches a narrow set of expensive-to-discover pipeline mistakes before
an adapter is released: malformed chat records, conflicting or leaked examples,
credential-shaped text, broken EOS handling, prompt tokens included in loss, and
answer-destroying truncation. It then gates paired base-versus-adapter scores
against explicit target and retention budgets.
I built SFTGuard to catch a narrow set of expensive-to-discover pipeline
mistakes before an adapter is released: malformed chat records, conflicting or
leaked examples, credential-shaped text, broken EOS handling, prompt tokens
included in loss, and answer-destroying truncation. It then gates paired
base-versus-adapter scores against explicit target and retention budgets.

It runs locally, emits content-addressed JSON, and returns `PASS`, `FAIL`, or
`ABSTAIN`. Missing evidence never becomes a silent pass.
Expand Down