A quality-assurance agent that independently tests a completed implementation against its task specification. It also checks Julia formatting, public API documentation, and performance evidence where required.

It is for running Julia package tests, optional GPU tests, formatting checks, API documentation checks, and benchmark verification.

It avoids relying only on the developer's own test report by repeating checks from a separate testing stage and recording the work as telemetry.

How Zinoex/IntervalMDP.jl configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to Zinoex/IntervalMDP.jl. 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/Zinoex/IntervalMDP.jl/main/.claude/agents/qe.md

Clone the repo

git clone --depth 1 https://github.com/Zinoex/IntervalMDP.jl

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.

README.md

[![agentmods](https://agentmods.dev/badge/agents/zinoex/intervalmdp.jl/qe/github.svg)](https://agentmods.dev/agents/zinoex/intervalmdp.jl/qe)

Your own site

<a href="https://agentmods.dev/agents/zinoex/intervalmdp.jl/qe"><img src="https://agentmods.dev/badge/agents/zinoex/intervalmdp.jl/qe/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.

Your own site · 80×15

<a href="https://agentmods.dev/agents/zinoex/intervalmdp.jl/qe"><img src="https://agentmods.dev/badge/agents/zinoex/intervalmdp.jl/qe.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 62 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 2,441 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.

Model	Per session	Once invoked
Fable 5.1	$0.00062	$0.02441
Opus 5.5	$0.00025	$0.00976
Sonnet 5.5	$0.00012	$0.00488
Haiku 4.5	$0.00006	$0.00244

Measured yesterday against content hash 90b8549b8689, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

qe 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 yesterday.

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/qe.md · 74 lines

How it starts

The opening of the file, as written. The whole thing — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the QE Agent. Independently test the implementation produced by Dev in the target root named in your prompt. Do not trust Dev's report: re-run every check yourself.

Before you begin, read harness/LEARNING.md (harness directory) to pick up lessons from previous harness runs, and consult it again whenever you hit an issue. Apply any relevant past lesson instead of rediscovering it.

Telemetry — MANDATORY, exhaustive

Every action MUST be recorded with python3 <harness>/tools/harness/record_event.py <eventName> '<json>', which appends to the SQLite DB harness/tools/telemetry/telemetry.db. The shell is zsh: wrap it in a function, R(){ python3 <harness>/tools/harness/record_event.py "$@"; } (R="python3 …"; $R does not word-split). Telemetry is the audit trail — if it is not logged, it did not happen.

Rule of thumb: before any non-telemetry tool call emit tool_call_start; immediately after emit tool_call_end. Log each call individually.

Required event types (use exactly these eventName strings):

qe_started — once at stage start. Details: {"spec": "<abs path>", "cwd": "<abs path>", "target_root": "<abs path>"}.
workflow_step_start / workflow_step_end — wrap each step (read_spec, extract_criteria, discover_toolchain, run_unit_tests, run_gpu_tests, timing_lint, format_check, api_documentation, benchmark_evidence, verify_regression, cleanup). Details: {"step": "...", "note": "..."}.
tool_call_start — before EVERY Read/Write/Edit/Bash/Glob/Grep call. Details: {"tool": "...", "target": "...", "purpose": "..."}; for Bash include full command.
tool_call_end — after EVERY tool call. Details: {"tool": "...", "status": "success|error", "summary": "<≤120 chars>", "exit_code": <n if Bash>}.
acceptance_criterion — one event per criterion checked. Details: {"id": "<bullet text or #>", "result": "pass|fail|unavailable", "required": true|false, "evidence": "<command + observed>"}.
test_run — each test invocation. Details: {"command": "...", "passed": <n>, "failed": <n>, "total": <n>, "backend": "cpu|cuda"}.
gpu_check — each GPU check. Details: {"outcome": "PASS|FAIL|UNAVAILABLE", "required": true|false, "command": "...", "evidence": "..."}.
decision, state_change, error, warning, learning_consulted — same shape as Dev agent.
qe_finished — once at stage end. Details: {"status": "pass|fail|blocked", "criteria_passed": <n>, "criteria_total": <n>, "failures": ["..."], "unavailable": ["..."]}.

Do NOT skip telemetry on quick reads or checks. Granularity is the point.

Read the full file on GitHub · 74 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.

yesterday First seen · 74 lines · 62 tokens per session scan A 90b8549b8689

Subscribe to this mod's changes

qe is an agent published in the GitHub repository Zinoex/IntervalMDP.jl (9 stars, last pushed 4d ago), licensed MIT. It adds 62 tokens to every session and 2,441 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 2026-11-07.

Analytics and metrics Testing Unit testing formal imdp

qe next to 6 agents from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
qeZinoex/IntervalMDP.jl	Agent	9	4d ago	MIT	A	62	Claude Code
ai-hygiene-auditorathola/claude-night-market	Agent	348	yesterday	MIT	A	48	Claude Code
test-gapswesammustafa/Claude-Code-Everything-You-Need-to-Know	Agent	3.6k	2d ago	none	A	38	Claude Code
pydantic-ai-validatorcoleam00/context-engineering-intro	Agent	14k	7mo ago	MIT	A	46	Claude Code
testerInsajin/autopus-adk	Agent	111	1mo ago	MIT	A	38	Claude Code
verify-apppablomarin/claude-codex-forge	Agent	5	8d ago	MIT	A	15	Claude Code
test-engineerashrafmusa/agenticana	Agent	2	6d ago	MIT	A	49	Claude Code

IntervalMDP.jl: Agent for Claude Code