Results for “agent-supervision”

Watch, audit, and fix another agent's work by reconstructing the original request, inspecting evidence, and reporting gaps.

Central status dashboard for the PA agent. Use when: owner asks 'what's the status', 'what are you working on', 'what's happening', or any status/overview question. Aggregates active tasks, open issues, monitored groups, pending follow-ups, and system health into one structured report.

Mental models and decision frameworks for building autonomous agents that continuously learn, explore, and accumulate skills in open-ended environments without human supervision

Agent skill for agent - invoke with $agent-agent

Agent skill for refinement - invoke with $agent-refinement

Agent skill for researcher - invoke with $agent-researcher

Analyzes repeated agent failures or noisy routing and proposes evidence-backed changes to instructions, skills, gates, or runbooks.

Agent skill for reviewer - invoke with $agent-reviewer

Manages GenAI tuning jobs in Agent Platform by listing, inspecting, or canceling ongoing model tuning jobs.

Run multiple AI agents against a product, SDK, or docs site to score setup friction, speed, efficiency, error recovery, and doc quality from a minimal prompt.

Check any AI agent codebase against the OWASP Agentic Security Initiative (ASI) Top 10 risks, scanning for controls and generating a compliance report.

Agent skill for architecture - invoke with $agent-architecture

`analysis-agent`/`task-agent`/`review-agent`: use when behavior needs deterministic seams for time, randomness, UUIDs, collaborators, or external I/O; skip when seams are adequate.

Use the built-in `Computer` sub-agent with `agent-desktop` for macOS desktop automation. Apply when a task needs application launching, accessibility snapshots, stable element refs, window focusing, semantic clicks/typing, or visual confirmation outside the browser sandbox.

Operate and extend barretlee/agent-pulse, the evidence-backed AI industry intelligence system: inspect source catalog and lifecycle, collect and normalize signals, bind evidence, cluster Events, evaluate system health, generate Scout hypotheses, export the privacy-safe public site, and verify release gates. Use when the user asks to run, configure, debug, extend, or explain Agent Pulse, its collectors, Control Room, narratives, Scout, or GitHub Pages output. Triggers on: agent-pulse, Agent Pulse, evidence-backed intelligence, source catalog, signal collection, Event clustering, source audit, Scout opportunity, public export, weekly brief, or AI industry intelligence pipeline.

Plan, execute, document, and retest authorized security assessments of AI agents and multi-agent workflows using safe adversarial cases, synthetic identities, canaries, and evidence-based findings. Use when defining red-team rules of engagement, assessing prompt injection or excessive agency, testing tool and identity boundaries, evaluating memory or cross-agent attacks, scoring a campaign, or verifying remediation in an approved environment.

`analysis-agent`/`task-agent`/`review-agent`: primary-Skill-selected for races, locks, optimistic conflicts, or worker overlap; never task owner; skip without concurrency impact.

`analysis-agent`/`task-agent`/`review-agent`: use for recurrence guards on known defects, incidents, or escaped failures; skip speculative risk without a prior failure mechanism.

Agent profile for inspect screenshots, viewports, layout overlaps, visual regressions, spacing, typography, and interaction states. Use when Codex needs a specialist agent perspective for planning, implementation, review, debugging, validation, or handoff in this domain.

Simulator.Company access-control specialist — who can view/modify/remove/sign/execute an object (actor, form, account, template, tree layer). Use when the user wants to share or unshare an object, grant or revoke permissions, list who has access, or bulk-share. Activate when the user says "share this with", "give access to", "grant permission", "revoke access", "who can see/edit this", "set permissions", "make read-only", "поділись з", "надай доступ", "забери доступ", "хто має доступ", "права доступу", "дай доступ", "открой доступ", "отзови доступ", "кто имеет доступ", "права доступа". Also activate when a tool fails because the caller has NO access to an object — "request access", "I can't see this actor", "access denied", "403", "no permission", "запроси доступ", "немає доступу", "немає прав", "відмовлено в доступі", "запроси доступ", "нет доступа", "доступ запрещён". For the object's own data use the domain skill (`simulator-actors` / `simulator-forms` / `simulator-finance`).

Agent profile for review API platform contracts, versioning, auth, rate limits, SDK clients, and developer experience. Use when Codex needs a specialist agent perspective for planning, implementation, review, debugging, validation, or handoff in this domain.

Discovers, installs, and manages AI agent skills from agentskill.sh, including searching, installing mid-session, scanning for security issues, and providing feedback.

Execute implementation plans by dispatching a fresh subagent per task, with task review after each and a whole-branch review at the end.

Decompose a task into parallel workstreams, assign agent ownership, run integration at dependency boundaries, and synthesize results.

Agent skill for planner - invoke with $agent-planner

Agent tooling for Motion Canvas — seek, screenshot, scene graph inspection, settings control, and rendering via HTTP API. Requires a browser with the editor open.

Rates an agent's own output on five axes — accuracy, completeness, clarity, actionability, conciseness — producing a structured scorecard with evidence and improvement suggestions.

Agent profile for art direction in 3D web design, product configurators, hero scenes, AR previews, and immersive visual interfaces.

Evaluate everything the PA agent manages — tasks, skills, PA network health, billing, calendar connections, and memory quality. Use when: owner asks for an evaluation, wants to know what's working and what isn't, or requests a performance report. Combines supervisor status with quality scoring.

Multi-agent adversarial verification with convergence loop. Two independent review agents must both pass before output ships.

Senior software engineer for story execution and code implementation. Use when the user asks to talk to Amelia or requests the developer agent.

Use this skill for SEO strategy, technical SEO prioritization, organic growth roadmaps, search risk review. Trigger when the task involves agent profile work related to SEO Director Agent, implementation, audits, debugging, strategy, or validation.

Use this skill for reproduction-first debugging, root cause isolation, minimal fixes, regression tests. Trigger when the task involves agent profile work related to Staff Debugger Agent, implementation, audits, debugging, strategy, or validation.

`analysis-agent`/`task-agent`/`review-agent`: use when uploads, object storage, streaming, MIME, scanning, access, retention, or cleanup changes; skip without file/storage impact.

Operate long-lived agent workloads with observability, security boundaries, and lifecycle management.