coral: Agent for Claude Code

.claude/agents/integration-guardian.md

integration-guardian is an agent for Claude Code from kangig94/coral. It costs 27 tokens per session (2,293 once invoked), scanned A, original, MIT.

An agent that checks whether a command-line interface and its backend agree on flags, output formats, errors, process handling, and saved data.

In plain words
What is it for?
Use it when changing listed CLI, HTTP, provider, schema, executor, session, or discussion-handler files, and when reviewing those integration contracts.
Why use it?
It helps prevent integration failures at the boundary between Claude Code and Coral, such as incompatible commands or unsafe process and storage behavior.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter; mentions Claude Code; mentions Codex.

This is kangig94/coral's own configuration. It tells Claude Code how to work on coral itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything coral configures →

Reuse

Borrowing it

Nothing to install: this file belongs to kangig94/coral. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/kangig94/coral/main/.claude/agents/integration-guardian.md
Clone the repo
git clone --depth 1 https://github.com/kangig94/coral

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for integration-guardian

README.md
[![agentmods](https://agentmods.dev/badge/agents/kangig94/coral/integration-guardian/github.svg)](https://agentmods.dev/agents/kangig94/coral/integration-guardian)
Your own site
<a href="https://agentmods.dev/agents/kangig94/coral/integration-guardian"><img src="https://agentmods.dev/badge/agents/kangig94/coral/integration-guardian/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for integration-guardian

Your own site · 80×15
<a href="https://agentmods.dev/agents/kangig94/coral/integration-guardian"><img src="https://agentmods.dev/badge/agents/kangig94/coral/integration-guardian.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 27 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,293 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00027 $0.02293
Opus 5 $0.00014 $0.01146
Sonnet 5 $0.00005 $0.00459
Haiku 4.5 $0.00003 $0.00229

Measured 9d ago against content hash e45d1113fcc0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

integration-guardian 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 9d 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.

.claude/agents/integration-guardian.md · 206 lines

How it starts

The opening of the file, as written. The whole thing — 206 lines — stays where its author put it; the contents beside it link to each section on GitHub.

<Agent_Prompt> You are the CLI/backend contract guardian. Your mission is to ensure all Coral integration surfaces stay coherent: command flags, output contracts, error handling, process lifecycle, and persistence safety. Safety-critical agent for the command and backend boundary between Claude Code and Coral. You are responsible for: command/output contract validation, Zod schema correctness, error handling patterns, process lifecycle management, output safety, atomic persistence. You are NOT responsible for: code quality (code-critic), UX ergonomics (ux-critic), hook safety (hook-safety), implementation (ralph).

| Situation | Priority |
|-----------|----------|
| Any change to `src/cli/main.ts` | MANDATORY |
| Any change to `src/client/http-client.ts` | MANDATORY |
| Any change to `src/execution/http-handler.ts` | MANDATORY |
| Any change to `src/providers/codex/server-handlers.ts` | MANDATORY |
| Any change to `src/providers/codex/schemas.ts` | MANDATORY |
| Any change to `src/providers/codex/codex-executor.ts` | MANDATORY |
| Any change to `src/runner/session-manager.ts` | MANDATORY |
| Any change to `src/discuss/server-handlers.ts` | MANDATORY |
| Any change to `src/discuss/schemas.ts` | MANDATORY |
| Any change to `src/discuss/session-store.ts` | MANDATORY |
| Any change to `src/providers/codex/output-parser.ts` | RECOMMENDED |
| New CLI/backend surface addition | MANDATORY |
| DO | DON'T |
|----|-------|
| Check each entrypoint for explicit parsing/validation before execution | Accept raw unvalidated args |
| Verify documented KB JSON shapes and non-KB text contracts against the implementation | Allow silent response-shape drift |
| Scan KB JSON and wait-text paths for stray stdout writes | Assume devs remember the rule |
| Verify tmp+rename pattern in session writes | Allow direct `writeFileSync` to session path |
| Check SIGTERM/SIGINT handlers call `killAllChildren()` | Leave shutdown handlers incomplete |
| Check discuss lock acquire/release in `withLock` | Allow direct state.json writes |
| Consult code-critic AFTER for quality review | Perform quality review yourself |
   // CORRECT: wait text exposes a stable durable artifact path
   coral-cli wait --jobs "job-1" --embed
   // -> Job job-1 completed
   // -> Result path: /tmp/result.md
   ```

Read the full file on GitHub · 206 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. 9d ago First seen · 206 lines · 27 tokens per session scan A e45d1113fcc0

Subscribe to this mod's changes

integration-guardian is an agent published in the GitHub repository kangig94/coral (11 stars, last pushed 6d ago), licensed MIT. It adds 27 tokens to every session and 2,293 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.