What it does
The agent audits what publication would actually expose, looking beyond the current files to the full commit history. It checks the author email on each commit, credential-like strings that can trigger GitHub push protection or vendor partner scanning, machine-specific paths, files committed before .gitignore covered them, and README claims that do not reproduce from a clean clone. Its stated focus is what leaves the project, not what an attacker could do once inside it.
It is meant for use before a repository becomes public: a first release, an internal project being open-sourced, or a private repo about to switch visibility. The publisher stresses that publication is hard to undo. A force-push does not recall copies a mirror or clone has already fetched, and a secret removed in a later commit still ships with every clone, which is why the whole history is checked.
The publisher's description does not say how the agent is invoked or what it needs in order to run.
Limitations and requirements
Scope is limited to what publication exposes. It is not a vulnerability audit and does not assess what an attacker could do from inside the system.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- AI Templates (aitmpl) first seen 13 Nov 2026 · last seen 18 Nov 2026 · its page on the source · figures reported by the source: aitmpl downloads 55
repo-publication-auditor
Use this agent before a repository becomes public — a first release, an internal project being open-sourced, or a private repo about to be flipped. It audits what publication actually exposes: the full commit history rather than the working tree, the author email on every commit, credential shapes that trip GitHub push protection and vendor partner scanning, machine-specific paths, files committed before .gitignore covered them, and README claims that do not reproduce from a clean clone. Distinct from a vulnerability audit — this is about what leaves the building, not what an attacker could do
As shown on www.aitmpl.com when we last read it