
argus
Persistent reviewed research and engineering runtime
- version
- 0.1.8
- GitHub stars
- 336
Install
/plugin install argus@argusRun it inside Claude Code after adding the marketplace with /plugin marketplace add lbx154/Argus.
Contents
Discovered from sourceSkills 3
Run With Argus
Use the installed Argus MCP tools. Do not imitate the Argus roles in the host session and do not write Argus state files directly.
- Call
argus_project_listwith the exact current work directory. - Reuse the most recently active exact-workdir project when one exists;
otherwise call
argus_project_createwith that directory and a concise name. - Call
argus_messagewith the user's complete objective. Manager owns chat versus task routing, lifetime, workflow, domain, and execution handoff. - Report the project ID and whether the turn was handled inline or dispatched.
- For dispatched work, stop the current turn. Do not create an unchanged polling
loop; use the
argus-statusSkill when status is requested.
If the MCP server is unavailable, run argus --doctor and report the exact
installation or backend failure. Do not silently replace Argus with an ad hoc
single-agent workflow.
Never claim completion from dispatch, daemon activity, loss, or intermediate artifacts. Argus completion comes from its persisted Reviewer/Manager state.
Inspect And Control Argus
- Resolve the project ID from the user's explicit ID or call
argus_project_listfor the exact work directory. - Call
argus_status. Report daemon state, active role, completed and pending work, pending operator questions, and continuous objective separately. - Call
argus_artifactswhen the user asks for outputs. Report only existing allowlisted artifacts returned by Argus. - Call
argus_doctoronly when status exposes a runtime/backend fault or the user requests diagnostics. - Send ordinary steering or answers through
argus_message; Manager remains the operator front door. - Call
argus_stoponly on an explicit stop request. Keepforce=falseunless the user explicitly requests a force stop after understanding interruption risk.
Do not count provider, transport, interrupted, or resource failures as task failures. Do not translate training loss or activity into a successful result.
Target-Disease Research
This workflow supports pharmaceutical research and portfolio decisions. It does not diagnose a patient, select treatment, or assert that a new drug was created.
- Require a target and disease. Capture aliases, population or subtype, date bounds, intervention class, exclusions, decision question, and output language when supplied.
- Stop if the request contains patient-identifying information and ask for a de-identified research question.
- Reuse only the project-resolution part of
argus-run: callargus_project_listfor the exact work directory, then reuse that project or callargus_project_create. Do not dispatch while resolving the project. - Call
argus_messageexactly once with an objective that asks Manager to select themedicalvertical (shipped by theargus-verticalspackage) and to produce:medical/evidence.jsonl,medical/evidence_matrix.csv,medical/target_disease_memo.md, andmedical/review.json. - Require source IDs and URLs, exact retrieval/query provenance, conflicts, failures, missing fields, and a separate infrastructure-failure count.
- Treat PubMed metadata as metadata, not full-text verification. Treat trial registration as trial existence/design, not evidence of efficacy. Qualify cross-trial comparisons unless population, endpoint, comparator, follow-up, treatment line, and data cutoff align.
- Return the project ID. Use
argus-statusfor later progress or artifact inspection instead of polling in the same turn.
Only describe a dossier as reviewed when Argus returns a Reviewer-certified state. Preserve negative and contradictory evidence.
MCP servers 1
Sponsors
Details
- Marketplace
- lbx154/Argus
- Source
- plugins/argus
- Category
- Productivity