kochab AGENTS.md

Standing instructions for agents and maintainers working on Kochab, a set of Claude skills for helping with job searches.

In plain words
What is it for?
Finding documentation, skills, packaged bundles, synthetic test data, release records, backlog items, and the rules to read before proposing changes.
Why use it?
It explains the project’s purpose, file layout, ethics, scope, and working rules so changes fit the product and do not expose personal job-search data.

Instructions file for CodexOpenCode

Install

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.

agentmods
npx agentmods add instructions/btmoriarty/kochab/agents-md
Clone the repo
git clone --depth 1 https://github.com/btmoriarty/kochab

Made for: Codex, OpenCode.

Per session 787 This file is loaded in full into every session.
When invoked 787 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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 $0.00787 $0.00787
Opus 5 $0.00394 $0.00394
Sonnet 5 $0.00157 $0.00157
Haiku 4.5 $0.00079 $0.00079

Measured 2d ago against content hash 1b8aec3375a1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

kochab AGENTS.md 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.

AGENTS.md · 41 lines

How it starts

The opening of the file, as written. The whole thing — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md — Kochab maintainer standing orders

These are the standing orders for any agent (or person) working in this repo. They mirror the Cowork project Instructions and travel with the code, so the rules hold in Claude Code and for anyone who forks Kochab. Read this, ETHICS.md, and ROADMAP.md before proposing changes.

What Kochab is

A job-search assistant shipped as a set of Claude skills. This repo is the product's source and maintainer workspace, not anyone's live job search. Personal data never lives here.

What lives where

  • README.md — what the product is and how to install it.
  • ETHICS.md — the non-negotiables. The reason the project exists.
  • ROADMAP.md — scope, version order, and what is permanently out of scope.
  • CHANGELOG.md — the record of what changed, per version.
  • skills/ — the source for each skill (a folder with SKILL.md at its root).
  • *.skill — packaged ZIP bundles, built from the folders in skills/.
  • docs/BACKLOG.md — candidate features and the need under each; where roadmap-interview intake lands.
  • docs/ — design brief, walkthrough, deck, getting-started, backlog, release checklist.
  • fixtures/ — synthetic test data. The only "user data" allowed in the repo.

Non-negotiables (from ETHICS.md; never violate, reject changes that would)

  1. The tool drafts; a human applies. No auto-submit, form-filling, or bulk submission, ever.
  2. Respect every site's terms: no scraping behind logins, no CAPTCHA or rate-limit evasion. Prefer public postings, official search, and the user pasting a job description in.
  3. Never fabricate. Resume and cover material is reorganized truth only, and obeys the user's honesty guardrails even when overstating would score better.
  4. Fit is shown as justified sub-scores with a named gap, never a bare number or percentage presented as objective.
  5. Report thin days honestly. Do not manufacture matches.

Working conventions

  • Build one version at a time. Respect items marked DECIDED in the roadmap. Surface OPEN ones for a decision instead of guessing. Treat "Deliberately out of scope" as a hard wall.
  • Every skill is a folder with SKILL.md at its root. Keep the description trigger-focused so it loads on the right tasks. Package a skill as a ZIP of its folder to install.
  • Test against fixtures/, never against real personal data. No profiles, resumes, trackers, or real employer names get committed here.
  • When behavior changes: update the relevant SKILL.md/references, add a CHANGELOG.md entry, repackage the .skill bundle (see docs/RELEASING.md), and say what changed and where.
  • Voice: plain and direct. No em dashes. No AI-cliche filler.
  • Em-dash convention (decided 2026-07-14): no em dashes in content the tool generates for a user (profiles, resumes, drafts, trackers, scan notes) or in templates the tool writes into a user's folder (for example the Application_Tracker.md header in references/pipeline.md), since those inherit the user's own voice rules; the synthetic fixtures follow the same rule. The repo's own narrative docs (README, walkthrough, CHANGELOG, ROADMAP) keep the author's existing em-dash style. When in doubt in generated content, use a hyphen.

Read the full file on GitHub · 41 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. 2d ago First seen · 41 lines · 787 tokens per session scan A 1b8aec3375a1

Subscribe to this mod's changes

kochab AGENTS.md is an instructions file published in the GitHub repository btmoriarty/kochab (1 stars, last pushed 1mo ago), licensed MIT. It adds 787 tokens to every session, about $0.0039 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.