agentleFS
landedjobslandedjobs/ai-job-hunt-os/skills/cold-outreach-writer/SKILL.md

cold-outreach-writer

Skill3 starsChanged 4 months agoMITView on GitHub

Use when the user wants to message a hiring manager, founder, engineer, or recruiter at a target company, ask someone for a referral or advice, follow up on an application, or verify whether a posted role is actively interviewing.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. jqaisystems/jqai-ai-skills★ 3same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository6 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When the real goal is checking a suspicious posting (high ghost risk from jd-decoder), the ask is simply: "Is this role actively …

    line 27
  2. Distinctive:

    Write three variants with different angles (product-hook, mutual-context, direct-and-brief), then recommend one and say why.

    line 31
  3. Distinctive:

    If the user is running a campaign, add a tracking table: persona, channel, message angle, proof asset, sent date, reply.

    line 38
  4. Distinctive:

    Never fabricate the connection point, inflate the proof line, or use flattery that pastes onto anyone ("love what you're building!").

    line 42
  5. Distinctive:

    No em dashes, no "I hope this finds you well," no "I'm reaching out because." Write like a person typing to a person.

    line 43

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Cold Outreach Writer

Write messages that busy people can answer, and set honest expectations about the odds. Do not quote universal response rates: outcomes vary sharply by role, seniority, channel, sender credibility, and what a source counts as a response. The durable lesson is narrower: a genuinely specific note gives the recipient a reason to engage, while generic volume creates no signal. Prefer a handful of researched messages over indiscriminate spraying, then measure the user's own sent, delivered, replied, and meeting-booked rates.

What you need

  1. Who to reach (name/role, or "help me pick the right person at company X").
  2. The actual goal: referral, direct consideration, a 15-minute conversation, intel, or verifying a posting is real.
  3. The user's one genuine connection point: a post the person wrote, a product decision the user noticed, shared background, or directly relevant work of the user's own. If there is no connection point, stop and build one first; a thoughtful comment on the person's recent post a few days before the DM changes the odds materially. A message with no specific hook is spam regardless of wording.

Picking the target (when asked)

The likely hiring manager beats the recruiter: recruiters run process, managers feel the pain of the unfilled seat. An engineer on the team is the best referral path. Founders reply surprisingly often under ~50 people; above that, target the team lead. Someone who posts publicly is 5x easier to open honestly, because there is something real to respond to.

Message anatomy (every message, every channel)

  1. Specific connection first: their post, their product, the exact problem their team is hiring for. Never open with the user's name and title; the reader does not care yet.
  2. One line of proof: the single most relevant thing the user has done, concrete, with a link if it exists. "Built an eval harness that caught 18% more citation failures" beats any adjective.
  3. One low-friction ask: advice beats referral on first contact ("would you point me at the right person?" / "worth 12 minutes?"). Never "any opportunities?" and never a resume attachment on first contact; it makes the note transactional.
  4. 50-75 words total. More than half the words should be about the recipient or their team, not the sender. On X, under 50 words.

When the real goal is checking a suspicious posting (high ghost risk from jd-decoder), the ask is simply: "Is this role actively interviewing?" Short, legitimate, and it converts uncertainty into information.

Output format

Write three variants with different angles (product-hook, mutual-context, direct-and-brief), then recommend one and say why. Channel register:

  • Email: include a subject line, specific and boring-good ("eval infra background, re: your agent reliability role"), never clever. Email has the best-measured odds.
  • LinkedIn DM: shortest, no formalities. Works best with visible shared context (alumni tie, mutual, recent post).
  • X DM: most casual; X is a relationship channel for people visibly active there, not a default application channel.

Then the follow-up plan: one nudge after 5-7 days that adds something new (a project link, a short teardown, a relevant update), never "just bumping this." Two messages total, ever. Include the fallback: who else to try, or apply through the front door.

If the user is running a campaign, add a tracking table: persona, channel, message angle, proof asset, sent date, reply. Reply rates by cohort tell you what to fix; untracked outreach cannot be learned from.

Hard rules

  • Never fabricate the connection point, inflate the proof line, or use flattery that pastes onto anyone ("love what you're building!").
  • No em dashes, no "I hope this finds you well," no "I'm reaching out because." Write like a person typing to a person.
  • Payload first: the reader should get something specific even if they never reply.
  • Set expectations honestly: most messages fail, including good ones. The strategy wins on the portfolio, not any single note.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.