test-run — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited test-run (Agent Skill) and scored it 96/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
A bulleted imperative like {match} tells the agent to never reveal, disclose, or mention something to the user. Used adversarially it can instruct the agent to hide its tool calls or lie about what it did — stripping the transparency a user relies on to trust the agent.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Manually trigger the daily or remote job search scripts outside their scheduled runs. Run all commands from the project root: C:/Users/mahas/Learnings/claude-job-agent.
Run daily_job_search.py. This reads List.xlsx, fetches hot jobs from LinkedIn, and sends the styled HTML digest email.
To force a full hot jobs refresh (re-fetch everything), delete daily_hot_jobs.json before running.
Run remote_search/remote_job_search.py.
Always add `--no-save` for test runs. Without it, previous_jobs.json gets overwritten and the NEW flag detection breaks for the real scheduled run. The Excel file still gets updated either way.
Only omit --no-save when running for real (i.e. as a substitute for the scheduled run).
config.py not found — copy config.template.py to config.py and fill in credentials.PermissionError on Excel — the script handles this with a temp copy fallback, but closing List.xlsx first is cleaner.EMAIL_CONFIG in config.py.~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.