agentmods.dev

dear-hiring-manager CLAUDE.md

A set of instructions for a Claude Code plugin that helps with job applications. Claude Code is the program that runs the plugin; the plugin reads job descriptions and user information while Playwright operates the web browser.

In plain words
What is it for?
Use it to onboard a user, extract and assess a job description, prepare application answers and a cover letter, fill forms through Playwright, and stop before submission for human approval.
Why use it?
It separates the agent's interpretation from browser actions and keeps the human in control before an application is submitted. It also explains where the plugin's commands, skills, templates, and local planning files belong.

Instructions file

108 1mo ago A scan 1,005 tokens original MIT
View the source file extrasmall0/dear-hiring-manager
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/extrasmall0/dear-hiring-manager/claude-md
Clone the repo
git clone --depth 1 https://github.com/extrasmall0/dear-hiring-manager
Per session 1,005 This file is loaded in full into every session.
When invoked 1,005 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 $0.01005 $0.01005
Opus 5 $0.00502 $0.00502
Sonnet 5 $0.00201 $0.00201
Haiku 4.5 $0.00101 $0.00101

Measured yesterday against content hash 09a22e410c97, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dear-hiring-manager CLAUDE.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 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.md · 61 lines

How it starts

The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — dear-hiring-manager

Guidance for any agent working in this repo. Also read PLAN.md, PRD.md, TODO.md if present — local planning docs, gitignored (not in fresh clones).

What this is

A Claude Code plugin that assists job applications: extract the JD, score fit, fill the form via Playwright, and stop before Submit for the human. Not a standalone agent — Claude Code is the runtime; this repo is the skill/knowledge layer. (Phase 3 later adds a real daemon.)

Architecture principle

AI understands; code + the browser execute. The LLM reads pages, fields, JD, matching, cover letters. Playwright does open/click/type/wait. Never make the LLM the hands.

Layout

  • .claude-plugin/plugin.json · marketplace.json — plugin + marketplace (dhm, hosted on GitHub)
  • .mcp.json — bundles Playwright MCP (@playwright/mcp)
  • commands/ — thin entry points (/dear-hiring-manager:onboard, :apply)
  • skills/onboarding/SKILL.md — interview → profile.md, parse resume, seed answers.md
  • skills/apply/SKILL.md — the assisted-apply procedure (the core logic)
  • templates/ — profile / answers / applications templates (copied to user data dir)
  • PLAN.md (phases) · PRD.md (design + traceability) · TODO.md (backlog) — local planning docs, gitignored (not shipped)

User data (never in repo)

~/.dear-hiring-manager/: profile.md, answers.md (append-only Q&A "RAG"), applications.md (tracker), resume.*. Flat Markdown — human-readable, no DB, no extra API keys.

Invariants (do not break)

  • Never click Submit. Human owns the last click.
  • Never solve/evade CAPTCHA. Hit one → stop, hand to human.
  • Portal accounts (Workday/iCIMS/etc.): agent may auto-create/login with the profile email + a unique generated password stored in the macOS Keychain (security CLI) — never a password in a file, never reused across portals. Email-verification + signup CAPTCHA/2FA → pause for the human.
  • Multi-page wizards: fill each page, click Next, stop only at the FINAL Submit.
  • Never auto-attest unconfirmed legal answers (non-compete/felony/etc.). Blank in profile → flag, don't fill. Legal defaults are written only after explicit onboarding confirmation.
  • Voluntary EEO blank → "prefer not to answer" (never flag, never block).
  • Work authorization is per role country — answer for the role's country, not the profile's raw yes/no (NL+EU auth + US role → authorized No, sponsorship Yes).
  • Fit score is the FIRST FILTER — below the profile's Minimum fit score to apply (default 50) or hard-ineligible → stop before filling anything.
  • Fill or flag every required field — never leave one silently blank.
  • answers.md is append-only — human edits win by being newer, never overwrite.
  • Tracker status lifecycle: in-progress (JD extracted) → skipped (gated) / filled (parked at Submit) → submitted.

Read the full file on GitHub · 61 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. yesterday First seen · 61 lines · 1,005 tokens per session scan A 09a22e410c97

Subscribe to this mod's changes

dear-hiring-manager CLAUDE.md is an instructions file published in the GitHub repository extrasmall0/dear-hiring-manager (108 stars, last pushed 1mo ago), licensed MIT. It adds 1,005 tokens to every session, about $0.0050 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 2026-08-30.

Browser automationPlanningResearchai-agentatsautomationclaude-codecover-letterhuman-in-the-loop
Related

Other instructions, from other repositories

Resume-Builder CLAUDE.md

Instructions for jananthan30/Resume-Builder, covering resume builder - project context, quick start, option 1: claude code custom commands (recommended), resume only - dual optimized and cover letter only.

jananthan30/Resume-Builder · 7,624 tokens

r3 AGENTS.md

Instructions for hyperlogue/r3, covering r3 — review. revise. resolve, architecture, module layout, domain model and review kinds & sources.

hyperlogue/r3 · 13,771 tokens

Resume-Builder AGENTS.md

Instructions for jananthan30/Resume-Builder, covering resume builder - codex context, codex plugin surface, native four-role resume team and operating rules.

jananthan30/Resume-Builder · 2,376 tokens

jobpilot CLAUDE.md

Instructions for suxrobGM/jobpilot, covering jobpilot, layout, commands, code style and area rules.

suxrobGM/jobpilot · 963 tokens

jobpilot AGENTS.md

Instructions for suxrobGM/jobpilot: Read CLAUDE.md first - architecture, layout, commands, and repo-wide code style.

suxrobGM/jobpilot · 133 tokens

resume-tailor-plugin CLAUDE.md

Instructions for olegvg/resume-tailor-plugin, covering claude.md, project overview, architecture, key concepts and external dependency.

olegvg/resume-tailor-plugin · 636 tokens