harness-scouter
Scores a Claude Code harness from the transcripts already on your machine: 6 stats of what your agent actually did, computed locally with no LLM calls. Then it runs a reproducibility gate over its own numbers and prints which ones failed and which it could not test at all.
Description supplied by the publisher; it is a claim, not verified evidence.
- Name
- github:sean5940/harness-scouter
- Listed in
- GitHub
- Publisher
- @sean5940
- Published
- 2026-10-10
- License
- MIT
- Evidence
- Listed only, not analyzed yet
Source: https://github.com/sean5940/harness-scouter
What a full analysis of harness-scouter checks
Not checked yet. About 60 tools are analyzed every hour; each check links to its source when it runs:
- Publisher identity: not checked yet
- Repository ownership: not checked yet
- Organization verification: not checked yet
- Artifact provenance: not checked yet
- Source availability: not checked yet
- Version evidence: not checked yet
- Known vulnerabilities: not checked yet
- Install scripts: not checked yet
- Maintenance: not checked yet
- License: not checked yet
Before you install it
- Open the repository and confirm the account is the one you expect: look-alike names are common.
- Read what it asks for: files, accounts, network access and credentials it needs.
- Prefer a version from a verified publisher when one exists.