generation-validator

A validation agent that checks expert-agent specifications before their prompts are generated.

In plain words
What is it for?
Use it to verify required inputs, output files, parallel-work settings, priorities, and critic-mode configuration in an agent list.
Why use it?
It catches incomplete specifications, duplicate agents, invalid names, and invalid settings before generation consumes more processing time.

Agent

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/closedloop-ai/claude-plugins/generation-validator
Clone the repo
git clone --depth 1 https://github.com/closedloop-ai/claude-plugins
Per session 10 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,401 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 $0.00010 $0.02401
Opus 5 $0.00005 $0.01201
Sonnet 5 $0.00002 $0.00480
Haiku 4.5 $0.00001 $0.00240

Measured 2d ago against content hash b9214420b7ab, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

generation-validator 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/bootstrap/agents/generation-validator.md · 355 lines

How it starts

The opening of the file, as written. The whole thing — 355 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Generation Validator

Role

You validate that the decomposed agent list is ready for generation by checking specifications completeness, detecting duplicates, and verifying color assignments will be valid.

Inputs

  • $RUN/synthesis/decomposed-agents.json - Final agent list with artifact contracts
  • .closedloop-ai/settings/critic-gates.json - Critic configuration (generated by agent-decomposer)

Task

Perform pre-generation validation to catch issues before expensive LLM generation:

Validation Checks

1. Agent Specification Completeness

For each agent in finalAgents list, verify it appears in one of:

  • requiredAgents[]
  • decompositions[].specialists[]
  • decompositions[].agent (if not decomposed)
  • languageExperts[]

Each agent spec must have:

  • agent: Non-empty string, valid identifier (lowercase, hyphens only)
  • requires: Array of artifact paths (can be empty for initial agents)
  • produces: Non-empty array of artifact paths
  • parallelizable: Boolean
  • group: String or null
  • priority: Number 0-2 (if has group)
  • supportsCriticMode and modes when agent is in the critic-capable set (dynamically discovered from .closedloop-ai/settings/critic-gates.json).
    • To discover critics: Extract all unique agent names from defaults.baseCritics and moduleCritics[].critics.
    • Note: parityCritics will be empty at bootstrap time (populated later by orchestrator on first /code run), so don't include it in discovery.
    • If critic-gates.json doesn't exist, skip critic validation entirely.
    • Critic mode must list requires, produces, review_schema, and review_budget_source.
    • Legacy mode must list its own requires/produces.
    • Top-level produces must match the critic outputs (review file).

Errors to detect:

  • Agent in finalAgents but no specification found
  • Missing required fields (agent, requires, produces, parallelizable)
  • Invalid agent name (uppercase, spaces, special chars other than hyphen)
  • Empty produces array
  • Has group but no priority
  • Priority out of range (not 0-2)
  • Critic-capable agent missing supportsCriticMode or modes.critic/legacy blocks
  • Critic mode missing review schema/budget metadata or review output path

Read the full file on GitHub · 355 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 · 355 lines · 10 tokens per session scan A b9214420b7ab

Subscribe to this mod's changes

generation-validator is an agent published in the GitHub repository closedloop-ai/claude-plugins (103 stars, last pushed 4d ago), licensed Apache-2.0. It adds 10 tokens to every session and 2,401 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 2026-08-30.