argus-agent v1.0.0

Source distribution and wheel for argus-agent v1.0.0.

Install directly from this repository at this tag:

pip install "argus-agent @ git+https://github.com/Inan15/Agent-Argus.git@v1.0.0"

Repository visibility, MEASURED 2026-08-29 by gh repo view Inan15/Agent-Argus --json visibility,isPrivate -> PUBLIC / isPrivate: false. What that buys a consumer, stated plainly: the pinned install resolves for anybody with no credential, and the GitHub Release and its attached packages are publicly downloadable. This SUPERSEDES the 2026-08-15 measurement, which read PRIVATE / isPrivate: true and said the pinned install cannot resolve for anybody — tag or no tag — without a read credential carried in the URL (git+https://<credential>@github.com/...) — accurate on its date, never re-run for fourteen days, and false for an unknown part of them. That is the whole hazard of this sentence: it is a dated measurement, not a standing claim, and re-running the command above before relying on it is the only thing that keeps it true.

The exit-code wire contract is UNCHANGED by this release: 0=RELEASE_READY, 1=no verdict produced, 2=NOT_READY_FOR_RELEASE, 3=INSUFFICIENT_COVERAGE. Exit 1 is reserved and is never a verdict — a run that produced no verdict made no statement about your code. See CHANGELOG.md for the full consumer contract.

CI evidence: run 33235322979 (ac1265e, 3/3 legs green) on audit-ci.yml covers the commit being released. Observed 2026-08-29 through the GitHub API.

SCOPE of that run, because a green run is evidence for what it EXECUTED and this one did not execute everything it carries. Each leg reported 1777 passed, 4 skipped. The run recorded the following as NOT EVALUATED rather than as passing, so the citation above does not reach them: (1) tests/test_installed_artifact.py (TC-ArgusAgent-RELEASE-001-25..-28) — the fresh-environment installed-artifact proof: every [project.scripts] console script, argus --help, a fixture audit run to a real verdict, and an MCP JSON-RPC exchange over stdio through the installed argus-mcp shim. All four SKIPPED on all three legs, each reporting the named E6 outcome NOT EVALUATED — uv is not on PATH, so the wheel could NOT be installed into a fresh environment and nothing about the INSTALLED distribution was checked. So the front-door claim of this release is held by LOCAL runs only, and this citation does not cover it. Provisioning uv on the CI runner is a tooling decision that has not been taken; it is filed OPEN and unscheduled as DF-12-9-B, owned by the Engineering Lead. Reading the citation as covering these would be the same class of overstatement as quoting a run id without the sha it covers.

This release makes no assurance claim about Argus itself. Argus's dogfood run is a self-audit and is not independent corroboration.

Beta: Argus's finding precision has not been independently validated. Its findings rest on the Argus dogfood corpus, a self-audit of this repository. Treat findings as a prompt to look, not a verdict. This notice is removed only when the >=80% precision gate is met; nothing else removes it.