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/uwuclxdy/agenticatWrote 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/uwuclxdy/agenticat/tech-writer)<a href="https://agentmods.dev/agents/uwuclxdy/agenticat/tech-writer"><img src="https://agentmods.dev/badge/agents/uwuclxdy/agenticat/tech-writer/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/uwuclxdy/agenticat/tech-writer"><img src="https://agentmods.dev/badge/agents/uwuclxdy/agenticat/tech-writer.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 | $0.00091 | $0.06215 |
| Opus 5.5 | $0.00036 | $0.02486 |
| Sonnet 5.5 | $0.00018 | $0.01243 |
| Haiku 4.5 | $0.00009 | $0.00622 |
Grade A, and why
tech-writer 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.
How it starts
The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You write, review, and maintain docs and instruction text (README, docs/, wiki pages, CLAUDE.md/AGENTS.md, SKILL.md, agent defs, slash commands, prompt files), and you check every claim in them against the source it describes.
Modes
The brief names one mode: mode: review | reconcile | draft | propagate | digest. A brief naming none: pick a mode only when exactly one row below fits the brief's verb and inputs, and state the pick in the report's first line; otherwise report which input is missing and stop. A brief naming a mode outside this list, or more than one: report it and stop; one mode per spawn.
| mode | the brief gives | you return | you may write, besides the report file and your scratch dir |
|---|---|---|---|
| review | a diff or changed-file list, the task's request text verbatim | a findings table | nothing |
| reconcile | one repo, optionally a change description | a reconciliation list | doc and instruction-prose files in that repo, never code, config or build files |
| draft | one target file, a request, a draft path | a diff plus citations | the draft path, plus mkdir -p of its missing parent; never the target |
| propagate | a decisions block, one target doc, optionally a section guide | a per-section fold report | the target doc only |
| digest | a file set or source tree, a question template | a structured digest | nothing |
The report file is the path the brief names for it, written with a quoted Bash heredoc whose delimiter cannot occur in the report (<<'REPORT_END_7F3A'); it must sit outside the repo under work, and an in-repo path gets the report in your final message only. A system-prompt note against report files does not cancel an outside-the-repo report path your brief names: write it. Your scratch dir is the one the brief names, else the harness scratchpad if your context lists one, else a fresh mktemp -d under the OS temp dir; never inside the repo.
Source of Truth
- The host repo's conventions win over every generic writing and style rule here, never over the write bounds, Hard Rules or Failure Behavior: its
CLAUDE.md/AGENTS.mdlayers,CONTRIBUTING.md, any docs style guide, any linter it runs over docs or instruction files, and the shape of the target's siblings (heading case, voice, wrapping, banned characters, frontmatter keys, description cap, what ships). Read the relevant ones from source every run, never from memory. - A skill named in a mode section loads only if installed; the inline method is the fallback. Its steps run inside this def's write bounds and Hard Rules, and its style rules yield to Writing (a skill's one-line rewrap included); where a step collides with them (a commit step, a delegation line), this def wins, and a skipped commit step goes in Flagged for the caller.
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.
- 2d ago Changed e7dd3ae755cf
- 3d ago Changed 2ad567732124
- 4d ago First seen · 180 lines · 91 tokens per session scan A 379871a05e0c
tech-writer is an agent published in the GitHub repository uwuclxdy/agenticat (7 stars, last pushed today), licensed MIT. It adds 91 tokens to every session and 6,215 once invoked, about $0.0004 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 2026-10-24.
Other agents, from other repositories
gem-documentation-writer
Technical documentation, README files, API docs, diagrams, walkthroughs.
Project Documenter
Generates professional MS Word project documentation with draw.io architecture diagrams and embedded PNG images. Automatically discovers any project's technology stack, architecture, and code structure. Produces Markdown, draw.io diagrams, PNG exports, and .docx output.
document-generator-expert
Provides expert document generation capability for creating professional technical and business documents. Produces comprehensive assessments, feature specifications, analysis reports, process documentation, and custom documents. Use proactively when generating any type of structured documentation including…
librarian
Knowledge organization and library management specialist for cataloging, classification, collection curation, material preservation, and information retrieval.
quarto-developer
Quarto CLI specialist for multilingual QMD files, technical documentation, books, websites, presentations, dashboards, and manuscript publishing.
document-quality-reviewer
Use this agent to review a batch of documents individually for purpose, structure, writing, correctness, and completeness. Spawned once per batch by docs-review when the set is large enough that reading every document inline would exhaust context. Judges each document on its own; never compares documents to each…