agent-operations-review

Reviews agent permission friction, tooling blockers, and outcome evidence after Work Blocks.

Updated Jun 12, 2026
One-click install
npx skills add https://github.com/oleyna80/agentic-sdlc-framework --skill agent-operations-review-oleyna80
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-operations-review
Source: https://github.com/oleyna80/agentic-sdlc-framework/tree/main/skills/agent-operations-review
Command: npx skills add https://github.com/oleyna80/agentic-sdlc-framework --skill agent-operations-review-oleyna80

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? After large Work Blocks, sprint closeouts, or multi-agent orchestration, teams lack a structured way to learn from permission friction, repeated approval prompts, sandbox failures, and subagent coordination issues. This Skill produces a local-only, read-only review with concrete recommendations for reducing friction without weakening safety gates. ## Core Features & Use Cases - Friction Analysis: Identifies repeated approval prompts, tool escalations, and sandbox/tooling failures using sanitized local evidence such as closeout summaries, git status, and verifier output. - Safety-Preserving Recommendations: Suggests allowlist candidates, skill updates, and workflow rule changes while explicitly preserving Hard Stops for DB, deploy, secrets, destructive git, and client communications. - Structured Output: Produces a report following the agent-operations-review template with sections for friction, blockers, outcome anchors, safe automation candidates, and next actions. - Use Case: After a multi-agent sprint where subagents repeatedly waited on permission approvals, run this review to decide which safe commands could be pre-allowed and which gates must remain unchanged. ## Quick Start Ask the agent to run an operations review of the last Work Block using the closeout summary, progress notes, and git status to identify permission friction and safe automation candidates.

Frequently Asked Questions about agent-operations-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review agent permission friction after a sprint?▼

Run this review after sprint closeout using the Work Block closeout summary, progress notes, git status, and verifier output as evidence. It identifies repeated approval prompts and tool escalations, then recommends allowlist candidates or workflow updates without changing permissions automatically.

What evidence does an agent operations review need?▼

It uses sanitized local-only evidence: Work Block closeout summaries, memory bank progress notes, git status and diff checks, verifier output, and outcome anchors like commit hashes and CI status. It never reads raw private transcripts, secrets, or full client messages.

Can this review change agent permissions automatically?▼

No. The review produces recommendations only and never modifies permissions, settings, or configurations. Any change to allowlists, workflow docs, or settings requires a separately approved Work Block with explicit owner sign-off.

When should I skip an agent operations review?▼

Skip it for small local changes with no permission friction, discussion-only turns, or Work Blocks where normal closeout already captured all blockers. Also skip if useful analysis would require reading raw private transcripts or secrets.

Which safety gates does the review never weaken?▼

It never recommends pre-allowing Hard Stops defined in AGENTS.md, and never weakens database, deploy, secrets, destructive git, or real client communication gates. Absence of negative evidence is treated as a signal to review, not proof of safety.