coach

A command that independently reviews the current implementation against requirements, such as a specification or requirements file.

In plain words
What is it for?
Run it with an optional requirements-file path, or let it look for common files such as requirements.md, SPEC.md, and TODO.md.
Why use it?
It checks each requirement and the build or tests, helping reveal incomplete work before the task is considered done.

Command

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 commands/agentic-insights/foundry/coach
Clone the repo
git clone --depth 1 https://github.com/Agentic-Insights/foundry
Per session 15 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 799 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 $0.00015 $0.00799
Opus 5 $0.00008 $0.00400
Sonnet 5 $0.00003 $0.00160
Haiku 4.5 $0.00002 $0.00080

Measured 2d ago against content hash 3b84559a6da1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

coach 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.

plugins/adversarial-coach/commands/coach.md · 119 lines

How it starts

The opening of the file, as written. The whole thing — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/coach - Adversarial Implementation Review

Validate the current implementation against requirements using an adversarial review approach from Block's g3 research.

Arguments

  • requirements-file (optional): Path to requirements document (e.g., SPEC.md, requirements.md)

Your Task

Step 1: Establish Requirements

If a file is specified:

  • Read the specified requirements file

If no file specified, search for:

  1. requirements.md, REQUIREMENTS.md
  2. SPEC.md, spec.md
  3. TODO.md with acceptance criteria
  4. Recent conversation context about what should be built

If nothing found:

  • Ask the user: "What requirements should I validate against?"

Step 2: Perform Adversarial Review

You are now the coach - review with fresh objectivity. Discard any knowledge of implementation decisions or shortcuts.

Systematic validation checklist:

  1. Requirements Compliance

    • Check each requirement item explicitly
    • Mark as ✅ implemented or ❌ missing/incomplete
  2. Build Verification

    # For the appropriate language/framework:
    # Rust: cargo build && cargo test
    # Python: uv run pytest or python -m pytest
    # Node: npm run build && npm test
    # Go: go build && go test ./...
    
  3. Common Gaps to Check

    • Authentication/authorization on all endpoints
    • HTTPS enforcement (if required)
    • Error handling for edge cases
    • Input validation at boundaries
    • Test coverage for error paths
    • Incomplete UI flows
  4. Functional Verification (if applicable)

    • Test actual API endpoints
    • Verify UI flows work end-to-end
    • Check that features actually function, not just compile

Step 3: Return Verdict

APPROVED (only if >95% complete):
IMPLEMENTATION_APPROVED

Validation summary:
- [Requirement 1]: ✅ Verified
- [Requirement 2]: ✅ Verified
- Build: ✅ Success
- Tests: ✅ Passing (N tests)
NEEDS WORK:
REQUIREMENTS COMPLIANCE:
- ✅ [Requirement]: Implemented correctly
- ❌ [Requirement]: [specific gap]

IMMEDIATE ACTIONS NEEDED:
1. [Specific, actionable fix]
2. [Specific, actionable fix]
3. [Specific, actionable fix]

Read the full file on GitHub · 119 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 · 119 lines · 15 tokens per session scan A 3b84559a6da1

Subscribe to this mod's changes

coach is a command published in the GitHub repository Agentic-Insights/foundry (5 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 15 tokens to every session and 799 once invoked, about $0.0001 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.