Argus Decision MCP server
io.github.commet/argus-decision-mcp·v2.1.0·1 starsPreserve a user-authorized judgment and append later answers. No model verdict or human score.
1 stars268 downloads/wk
Reviews
Write oneNobody has reviewed Argus Decision yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
Argus Decision tools (15)
write = sends, deletes, buys or postsRead from the package source without running it. The installed server may list more.
argus_amendAdjust an open or not-yet-due sealed decision's predicate or check-by date. Refused once the check-by date has arrived (no moving the goalpost after the fact).
argus_check_inReturn decision contracts whose check-by date has arrived (and optionally upcoming ones). A return nudge \u2014 reads and routes to argus_settle. If nothing is due, it says so and stops; it does not manufacture engagement.
argus_configRead or update non-spine settings (locale, boss, team, archive). Passing only argus_dir reads; passing fields merges and writes. There is no setting that turns off falsifiability, seal-before-settle, or honest provenance \u2014 the spine is not configurable.
argus_dismissClose a decision without settling it \u2014 the user moved on, decided elsewhere, or it became irrelevant. Terminal; not reopened.
argus_initInitialize the .argus directory (sessions, ledger, config, privacy .gitignore) and bind it for resource reads. Call once before other tools. Safe to call again.
argus_open_decisionOpen a consequential decision. Runs a fire-or-not restraint gate FIRST; if it fires, surfaces at most one neutral crux question and a "leave as is" option. Never a fork, never a verdict, never a lean. On flat/low-stakes/reversible/closed decisions it returns restraint.
argus_patternsRead active decisions, receipts, timelines, and recurring patterns. Read-only; history is context, never a verdict.
argus_planAttach the execution plan the user adopted to an open decision: ordered steps with optional check dates. Dated steps come back through argus_check_in on their date. Call ONLY after the user agreed to the plan in conversation; leave unknowns as open_questions instead of inventing steps.
argus_premisesRecord and maintain the premises a decision rests on \u2014 the facts and open questions behind it. op=add records them (echoed back in full so the user sees what was written); op=amend lets the user correct one (edits are the signal; provenance is preserved); op=resolve closes an open question in the user's own words. Premises lock once the check-by date arrives. A load-bearing external premise i
argus_recallRead your own decision history: a single receipt, open records, or a neutral chronology. Read-only. Answers are not aggregated into scores or performance claims.
argus_recheckRe-check one premise of a decision against reality. The host researches the fact and passes the finding (with provenance); the tool records it and decides drift mechanically \u2014 explicit numbers compare numerically, text premises take the host's asserted `changed`. First re-check records the baseline and never alerts. If the fact drifted, the response says so and returns the handle \u2014 wheth
argus_sealSeal a falsifiable prediction (predicate + check-by date). Locks it immediately; no prior capture step is required, and when the user says to seal, do it without re-asking. Seal each decision the user names, one call per decision. Captures the seal-time Judgment Receipt fields. Refuses an empty/non-falsifiable predicate or a non-future date. On success, show the short `surface` line as the confirm
argus_settingsRead or update language, reminders, and explicit sync settings. Argus initializes the current project on first use.
argus_settleSettle a sealed decision against reality and issue a Judgment Receipt with zero AI verdict. Hard-errors if there is no prior seal. The outcome is the user's \u2014 recorded, never inferred. On the ledger's FIRST completed settle, `surface` contains the full then-vs-now receipt \u2014 show it verbatim; it is the user's first payoff. Later settles return the short line and keep the receipt in data.r
argus_syncPull your Argus account receipts into the terminal \u2014 live judgments and what is due. Returns: receipts (id, local_id, settle_path, title, state, next_check_by, due, open_predicates) with due items first, plus total/due/has_more. Account ids carry an mcp_ prefix for terminal-sealed judgments: settle those with argus_settle using local_id (NOT the account id). Web-sealed judgments (local_id nul
Public scan report
scanner v0.1.10 · 2026-11-04 · same rubric, same numbers if you re-run it
- Code scan2 source files scanned20/25
- –Live reliabilityno gateway calls yet and no remote to proben/a
- –Tool poisoningtools not inspected (local package is not executed); not countedn/a
- Auth qualitystatic API keys via environment variables6/15
- Maintenancelast push 14 days ago15/15
- Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Findings (1)
- mediumHeavily hex-escaped string (obfuscation)
obf.hex-densitydist/index.js: …dict_line: "AI VERDICT ON THIS DECISION \xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7\xB7 NONE", closing: "The model never…
How Argus Decision compares
Against every graded AI and agents MCP server, from our own scans and uptime checks. Updated after each scan run.
- Safety score
- 75/100
- higher than 45% of 4,460 AI and agents servers
- Tools
- 15
- the typical AI and agents server has 6
- Tools that change things
- 0%
- typical: 0%. Fewer write tools means less an agent can break
- GitHub stars
- 1
- more than 50% of 2,711 AI and agents servers
- Downloads a week
- 268
- more than 78% of 1,538 AI and agents servers with a package
What its tools let an agent do: 15 read
Release history: 2 versions in the registry since 2026-08-12, latest 2026-10-21, typically one every 70 days.
Install Argus Decision in Claude Code, Cursor or VS Code
Runs npx -y argus-decision-mcp on your machine. Read the scan report first; the gateway never runs local packages.
claude mcp add argus-decision-mcp -- npx -y argus-decision-mcp
What the publisher says
From the Argus Decision repository's README, as published. We do not edit it. Read it on GitHub
Odysseus bound himself before the Sirens sang. Argus binds a decision to its premises and return condition before fluency — or hindsight — can rewrite them.
Argus
Keeping Judgment Human. Turn a decision into movement. Let reality shape the next call.
Web app · Install · What it does · The record · License · 한국어
Argus turns a stuck decision into one observable next move, keeps the words and assumptions you actually adopted, and brings the record back when reality can answer. The durable output is not another AI answer: it is a closed, append-only loop that can improve the next call without scoring the person.
Why "Argus"?
The mast above and the name Argus carry the two halves of the product. Odysseus bound himself before the Sirens could overtake his judgment. His hound, Argus, recognized him twenty years later beneath a beggar's rags. One holds a choice steady against persuasion; the other preserves memory, recognition, and fidelity beyond time and appearances.
That's the whole product. Work with an AI long enough and its smooth voice quietly takes over your memory of why you decided. Argus is the dog at the door. It remembers what you believed when you decided, notices the hidden premise under the fluent answer, keeps watch on the signals you chose, and returns first when reality has an answer.
It never decides for you, and it never grades the person. Reality can answer a claim; only you can answer what happened to your commitment or standard.
⚡ Try it in 30 seconds
Three doors, one loop. Start with whichever fits.
New and not sure? The web app is the durable Decision Desk and needs zero setup. Want Argus inside existing AI work? Use the MCP server or portable Agent Plugin. Full setup + tool list in argus-mcp/README.md and argus-plugin-v2/README.md.
Shortened. The full README is on GitHub.
Nothing above is checked by us. What we check is on the safety report.
Argus Decision: common questions
- Is Argus Decision MCP server safe?
- Mostly: it is graded B (75/100). Read the Argus Decision safety report
- How do I install Argus Decision?
- It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
- Does Argus Decision need an API key?
- Yes. The registry entry asks for
ARGUS_TOKEN. - Is Argus Decision maintained?
- The last commit was 14 days ago (2026-10-21). The latest release is v2.1.0.
- What can I use instead of Argus Decision?
- Servers from other publishers that do the same job: CHAP Coordinator MCP server, Decionis Commerce Gate MCP server and three.ws Provenance MCP server.
- How does Argus Decision compare with other AI and agents MCP servers?
- safety score 75/100 (higher than 45% of 4,460 AI and agents servers); tools 15 (the typical AI and agents server has 6); tools that change things 0% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.
Alternatives to Argus Decision
Same job from other publishers: the closest match first, then the best rated.
- CHAP CoordinatorAuditable records of human decisions over AI agent work. Approvals, edits, overrides, escalations.not reviewedEstablishedA
- Decionis Commerce Gate MCPCommerce preflights, D365 authorization; opt-in agent writes require authorized backend connectors.not reviewedEstablishedA
- three.ws ProvenanceAppend-only, signed, on-chain-verifiable agent action log — record and audit what agents did.not reviewedEstablishedB
Tw Legal RagCitation-guarded retrieval over 22M Taiwan court judgments and administrative interpretationsnot reviewedEstablishedA