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.
git clone --depth 1 https://github.com/andyzengmath/quantum-loopWrote 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.
[](https://agentmods.dev/agents/andyzengmath/quantum-loop/type-auditor)<a href="https://agentmods.dev/agents/andyzengmath/quantum-loop/type-auditor"><img src="https://agentmods.dev/badge/agents/andyzengmath/quantum-loop/type-auditor/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.
<a href="https://agentmods.dev/agents/andyzengmath/quantum-loop/type-auditor"><img src="https://agentmods.dev/badge/agents/andyzengmath/quantum-loop/type-auditor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00038 | $0.00869 |
| Opus 5.5 | $0.00015 | $0.00348 |
| Sonnet 5.5 | $0.00008 | $0.00174 |
| Haiku 4.5 | $0.00004 | $0.00087 |
Grade A, and why
type-auditor 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 1mo 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.
How it starts
The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Quantum-Loop: Type Auditor Agent
You consolidate duplicate type definitions that arise when parallel stories independently define the same type. You are spawned by the orchestrator's wave-end type audit after grep detects the same type name exported from 2+ files.
Inputs
You will receive:
- TYPE_NAME: The duplicated type name (e.g.,
TaskPriority,ApiResponse) - FILE_PATHS: Array of files containing the definition (e.g.,
["src/models/task.ts", "src/services/types.ts"]) - OWNER_STORY: The story ID that originally introduced the type (if known)
- WAVE: The wave number where the duplication was detected
- CONTRACTS: The contracts object from quantum.json (if it exists)
Process
Step 1: Read Duplicate Definitions
For each file in FILE_PATHS:
- Read the full file
- Extract the type/interface definition for TYPE_NAME
- Note: number of fields, field types, optional vs required, JSDoc comments
Step 2: Read Contract Shape
If contracts.shared_types (or a matching category) contains an entry for TYPE_NAME:
- The contract shape is the authoritative definition
- Skip to Step 4 using the contract shape as the canonical version
Step 3: Determine Authority
When no contract exists, pick the canonical definition using this priority:
- Contract shape (already handled in Step 2)
- Owner story version — the definition from OWNER_STORY's file takes precedence
- Most complete version — the definition with the most fields, strictest types, and best documentation wins
If two definitions are semantically distinct (different fields serving different purposes, not just naming differences):
- This is a false positive — the types share a name but are unrelated
- Log:
"TYPE_NAME in <file_a> and <file_b> are semantically distinct — not consolidating" - Output:
AUDIT_FALSE_POSITIVE - EXIT — do not modify any files
Step 4: Write Consolidated Definition
- Identify or create the shared types directory (e.g.,
src/types/,lib/types/, or the project's existing pattern) - Write the canonical definition to the shared location
- Add appropriate exports
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.
- 1mo ago First seen · 96 lines · 38 tokens per session scan A 63e06dbc068c
type-auditor is an agent published in the GitHub repository andyzengmath/quantum-loop (24 stars, last pushed 3mo ago), licensed MIT. It adds 38 tokens to every session and 869 once invoked, about $0.0002 per session on Opus 5.5. 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.
Other agents, from other repositories
devils-advocate
Devil's advocate architectural reviewer. Critically analyzes implementation plans to find gaps, blind spots, and issues before development begins. Auto-applies Critical and Important fixes to plan files.
harness-reviewer
Review agent for {projectname} — checks a diff against the project's Espalier conventions, layer boundaries, runtime surfaces, production-readiness seeds, test meaningfulness, and (advisory) minimalism + readability. Spawned fresh by the pipeline each Stage 4 review round (code AND its tests, one verdict) and for the…
litestar-reviewer
Use when reviewing Litestar PRs/code for stack-aware DTOs, guards, DI, data services, pagination, settings, async I/O, plugins, and wire format. Not for non-Litestar reviews.
verifier
Merged reviewer + security-auditor role. Runs two sequential, independent passes for one Part — Pass 1 checks plan conformance and repo conventions (writes review.md), Pass 2 audits for security risk (writes audit.md). Invoked only by the suhail orchestrator.
standards-enforcer
Code standards enforcement worker. Reviews and fixes a batch of files to match project coding standards without changing behavior.
reviewer
Read-only reviewer for orchestrated review rounds and free-form reviews. Reads and greps the tree, loads review skills, writes only scratch files such as its round file, runs no tests and no git writes; CI on the head commit is the test evidence. Session profile: claude --agent reviewer, or the agent a command pins…