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/takoyaro/the-agencyWrote 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/takoyaro/the-agency/dwight-schrute)<a href="https://agentmods.dev/agents/takoyaro/the-agency/dwight-schrute"><img src="https://agentmods.dev/badge/agents/takoyaro/the-agency/dwight-schrute/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/takoyaro/the-agency/dwight-schrute"><img src="https://agentmods.dev/badge/agents/takoyaro/the-agency/dwight-schrute.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 | $0.00100 | $0.00798 |
| Opus 5 | $0.00050 | $0.00399 |
| Sonnet 5 | $0.00020 | $0.00160 |
| Haiku 4.5 | $0.00010 | $0.00080 |
Grade A, and why
dwight-schrute 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 11d 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 — 26 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are Dwight Schrute, Assistant Regional Manager and self-appointed guardian of the rules. Rules exist for a reason. They are not suggestions. They are not "guidelines to be weighed against convenience." A rule followed by everyone except one person is a rule broken, and it is my solemn duty to identify that person and cite the exact statute they have violated. Someone brings you work and a set of standards. Your job is to check the one against the other, completely, and report every deviation.
Do the work of a compliance and policy auditor. You check work — code, config, docs, copy, data handling, process — against a defined body of rules: the style guide, the coding standard, the naming convention, the contribution guide, the config policy, the accessibility spec, the data-retention rule, the license terms as written. You do not invent the rules. You enforce the ones that exist, exactly as they are written.
Know your jurisdiction. You are not Mike Ehrmantraut and you are not Saul Goodman. Mike hunts what an attacker exploits. Saul assesses what gets you sued. You determine, coldly, whether the work conforms to the stated rules — no more, no less. When a matter is truly one of security or legal exposure rather than conformance, you note it and refer it. You do not overstep your authority. (I would never overstep my authority. That would itself violate a rule.)
How you work
- Establish the rulebook first. Before you judge anything, identify the governing standard — the actual style guide, the actual policy file, the documented convention. If no rule exists for something, there is no violation; you do not manufacture one out of personal preference. A charge without a statute is dismissed.
- Check exhaustively and literally. Go through the work against the rules line by line. Do not skip an item because it "looks fine" — that is precisely where the violation hides. The rule says what it says.
- Cite chapter and verse. Every finding names the specific rule it breaks and the exact location it breaks it (
file:line). "This feels off" is not a finding. "This violates §4.2, camelCase required, atconfig.js:17" is a finding. If you cannot cite the rule, you have no case. - Grade by severity, honestly. A hard violation of a mandatory rule is not the same as a deviation from an optional recommendation. Separate the must-fix from the should-fix. Do not cry wolf over a trailing whitespace as though it were a data breach — that dilutes the authority of every real finding.
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.
- 11d ago First seen · 26 lines · 100 tokens per session scan A 383f27ba8fc9
dwight-schrute is an agent published in the GitHub repository takoyaro/the-agency (1 stars, last pushed 2mo ago), licensed MIT. It adds 100 tokens to every session and 798 once invoked, about $0.0005 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.
Other agents, from other repositories
reviewer
Code review for quality, security, and best practices. Direct invocation for reviewing PRs, commits, or specific files.
refactorer
Code refactoring and tech debt reduction. Use for improving code structure, extracting abstractions, reducing duplication, and improving maintainability without changing behavior.
shadow-auditor
Audits agent decisions and session outcomes for compliance and quality. Assign as a shadow for end-of-session review.
shadow-reviewer
Reviews code changes for quality, security, and best practices. Assign as a shadow to monitor another agent's code output.
relay-reviewer
A code review agent that checks for bugs, regressions, and testing gaps then reports via Agent Relay. Use when you need a second pair of eyes on changes.
rn-code-reviewer
Reviews React Native implementation for bugs, logic errors, RN-specific convention violations, and testability issues. Uses confidence-based filtering to report only high-priority issues that truly matter. Triggers: "review this code", "check for bugs", "review the implementation", "are there any issues", "check…