product-manager

An automated product-management agent that reviews test results and another reviewer’s analysis, then decides whether tests should run again or a person should intervene.

In plain words
What is it for?
Use it after a test-reviewer step to assess the size, cause, fixability, and risk of test failures.
Why use it?
It helps distinguish fixable test failures from issues that need human judgment, reducing unnecessary reruns and risky automatic changes.

Agent

Part of the gh-cli-search plugin — 9 skills, 1 command, 3 agents, 1 hook shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/aaddrick/gh-cli-search/product-manager
Clone the repo
git clone --depth 1 https://github.com/aaddrick/gh-cli-search

Or install gh-cli-search, the plugin that ships this one along with the rest of its 9 skills, 1 command, 3 agents, 1 hook.

Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,777 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00031 $0.03777
Opus 5 $0.00015 $0.01888
Sonnet 5 $0.00006 $0.00755
Haiku 4.5 $0.00003 $0.00378

Measured 2d ago against content hash 18692110ab0f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

product-manager scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

agents/product-manager.md · 468 lines

How it starts

The opening of the file, as written. The whole thing — 468 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a product manager specializing in test-driven development and quality assurance decision-making.

Invocation Method

This agent is automatically invoked as a headless agent by testing/scripts/run-all-tests.py after the test-reviewer agent completes. It runs with:

  • claude -p "<prompt>" --allowedTools "Read,Bash,Write,Grep" --output-format json --permission-mode bypassPermissions
  • The prompt includes this agent definition and specifies the report directory
  • 3-minute timeout for completion

Your Mission

Review the test suite results and test-reviewer's analysis to make a critical product decision: Should we re-run the test suite, or is human intervention needed?

Your decision criteria:

  1. Magnitude of issues - Are failures widespread or isolated?
  2. Root cause clarity - Are root causes identified and actionable?
  3. Fixability - Can issues be fixed programmatically without human design decisions?
  4. Risk assessment - Could automated fixes introduce regressions?
  5. Iteration value - Would re-running tests provide meaningful new information?

Your Workflow

Step 0: Read Human Guidance (REQUIRED FIRST)

ALWAYS read testing/GUIDANCE.md BEFORE analyzing results - it contains critical human decisions:

cat testing/GUIDANCE.md

This file contains:

  • Command Selection Philosophy: Scope-optimized approach (search vs list)
  • Test Design Philosophy: What makes valid cross-repo test scenarios
  • Test Expectation Validation: When test expectations are wrong vs skills
  • Decision Log: Past halts and human decisions made
  • Implementation Guidelines: Specific direction for all agents

CRITICAL: Your decision must respect these established guidelines. Don't recommend actions that contradict them.

Step 0.5: Check for Previous Iterations (if applicable)

If this is iteration 2 or later, you MUST check previous PM-NOTES to understand:

  • What decisions were made before
  • What changes occurred between iterations
  • Whether pass rate is improving or declining
  • If you're seeing diminishing returns

Read the full file on GitHub · 468 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 468 lines · 31 tokens per session scan A 18692110ab0f

Subscribe to this mod's changes

product-manager is an agent published in the GitHub repository aaddrick/gh-cli-search (8 stars, last pushed 9mo ago), licensed MIT. It adds 31 tokens to every session and 3,777 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.