status

boshu2/agentops

Report observable AgentOps evidence: intent/verdict counts, recency, and source health.

What is status?

Displays a read-only status snapshot of AgentOps evidence stored in `.agents/ao`, reporting only verifiable facts from disk artifacts. Use this to check intent and verdict counts, evidence recency, and data integrity without modifying state.

Reports intent and verdict artifact counts from `.agents/ao/intents/sha256` and `.agents/ao/verdicts/sha256`
Determines evidence recency by identifying the newest stored artifact type
Detects and discloses corrupt or unavailable evidence sources
Explicitly labels checked vs. not_checked observations
Keeps distinct state classes separate (tracker, Git, factory, checks, validation)

How to install status

npx skills add https://github.com/boshu2/agentops --skill status

Prerequisites

AgentOps (ao) must be installed
`.agents/ao` directory must exist with intent and verdict artifacts

Claude Code

Cursor

Windsurf

Cline

How to use status

1.Trigger the skill with 'status' or 'show AgentOps status'
2.The skill reads `.agents/ao/intents/sha256` and `.agents/ao/verdicts/sha256` directories
3.Review the reported counts, recency conclusion, and any disclosed unavailable sources
4.Note which observations are marked `checked` vs. `not_checked`

Use cases

Good for

Check AgentOps evidence health during a session without triggering work selection
Verify intent and verdict artifact counts are consistent and accessible
Identify which evidence type is most recent (intent vs. verdict)
Diagnose missing or corrupted evidence in `.agents/ao` storage
Audit what AgentOps state is observable on disk at a given moment

Who it's for

SRE teams monitoring agent execution
Developers debugging AgentOps evidence storage
Teams using DORA metrics and observability practices
Anyone needing a trustworthy snapshot of local agent state

status FAQ

What does 'evidence recency' mean?

It identifies which artifact type (intent or verdict) is newest on disk, indicating the most recent AgentOps activity without inferring runtime phase.

Why does the report explicitly say 'not_checked'?

Runtime phase, execution time, tool activity, and remaining work cannot be determined from stored artifacts alone; disclosing gaps prevents false conclusions.

Can this skill modify or repair AgentOps state?

No. Status is read-only and never changes state, repairs records, or assigns work.

What counts as 'observable' evidence?

Only artifacts that exist on disk right now: validated intent and verdict files in `.agents/ao`, plus deterministic check results and source availability.

Why keep state classes separate in the report?

Tracker, Git, factory, and validation are independent authorities; blending them into one 'health' claim masks which specific system has an issue.

Full instructions (SKILL.md)

Source of truth, from boshu2/agentops.

name: status description: 'Report observable AgentOps evidence without selecting work. Triggers: "status", "show AgentOps status".' practices: [dora-metrics, sre] hexagonal_role: driving-adapter consumes: [] produces: [stdout] context_rel: [] skill_api_version: 1 allowed-tools: Read, Grep, Glob, Bash model: haiku context: window: inherit intent: {mode: none} metadata: capabilities: [status] effects: [read_filesystem, read_clock] canonical_status: canonical disposition: keep_specialist graph_root: true tier: session dependencies: [] output_contract: read-only status snapshot

Status

A status snapshot is trustworthy exactly when every line traces to an artifact that exists on disk right now; the first inferred line turns the report into a guess wearing a report's clothes.

Prompt

Show AgentOps status for fleet-router (ao is installed). Report only observable .agents/ao evidence: intent and verdict counts, evidence recency, and any corrupt or unavailable sources. Disclose checked and not_checked explicitly.

It's working if

The report cites ao status output directly, never a paraphrase of what it might show.
Intent and verdict counts trace to .agents/ao/intents/sha256 and .agents/ao/verdicts/sha256, never an inferred count.
Runtime phase and tool-call activity get labeled not_checked unless the caller supplied that source separately.
Tracker, Git, and factory state stay in their own reported lines rather than one blended health claim.

Contract

Report only observable local facts: available intent and verdict artifacts and their counts; deterministic check results; evidence recency; and unavailable or corrupt sources. The canonical durable stores are .agents/ao/intents/sha256 and .agents/ao/verdicts/sha256. Subject manifests are caller-supplied: report them only when the caller names their location, otherwise disclose them as not_checked. When .agents/ao evidence exists, report which stored artifact kind is newest and label that conclusion as evidence recency, not runtime phase or process activity. The ao status snapshot emits the two counts plus that newest-artifact recency conclusion, not a per-artifact digest or timestamp listing; report a specific artifact's digest or timestamp only when the caller asks about a named artifact.

Always disclose checked and not_checked. Runtime phase, execution elapsed time, tool-call activity, and remaining work are not_checked unless a caller provides a separate authoritative source for them.

ao status is the evidence-store view. It validates content-addressed artifact names and content before counting them, reports corrupt and unavailable entries, and shows only intent/verdict counts plus evidence recency. Retired legacy surfaces (session indexes, provenance summaries, knowledge-health signals) are not aggregated into this command.

Distinct state classes stay distinct in every status report: tracker state belongs to the tracker, Git state to Git, factory or runtime state to the selected factory's own doors, deterministic-check results to the executable that produced them, and semantic validation to a fresh verdict. Report each from its own authority, never merged into one blended "health" claim — factory-complete, checks-green, and AgentOps-PASS are three different facts.

Status does not inspect work queues, assign priority, claim work, infer a next action, repair records, govern retries, or change any state. Optional Git or tracker metadata may be displayed only when the caller supplies it; absence cannot change the report interpretation.

Named failure mode — recency-as-activity: reading "newest artifact is a verdict" as "validation is running", which invents a runtime phase from a timestamp.

Anti-pattern: filling not_checked gaps with plausible narrative so the snapshot feels complete. Corrective: report the gap as a gap; an honest hole outranks a smooth story.

Return the snapshot and stop.