agentsclimarketplace
Sign in

Recruiter reply

Skill evnm/claude-skills/recruiter-reply

A collection of personal Claude Code skillsFrom the repository description

Install
npx -y skills add evnm/claude-skills --skill recruiter-reply

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

  • skips confirmationTells the agent to proceed without asking first, 1 time: "Auto-skip it without prompting".
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
  • runs commandsInstructs the agent to run 5 commands, including `~/.claude/skills/recruiter-reply/.venv/bin/python ~/.claude/skills/recruiter-reply/gmail_helper.py fetch` and 4 more.

SKILL.md

5.3 KB, ~1.3k tokens by cl100k_base, as published. Nobody here has run it

recruiter-reply

Scan unread emails labeled "recruitment" for recruiter pitches, let the user choose a response for each, send replies, and update labels/read status.

Steps

1. Fetch labeled emails

~/.claude/skills/recruiter-reply/.venv/bin/python ~/.claude/skills/recruiter-reply/gmail_helper.py fetch

This prints a JSON array of unread emails with the "recruitment" label, with fields: gmail_message_id, rfc_message_id, thread_id, from_name, from_email, subject, body_text, date. The thread_id can be passed to fetch-thread to retrieve all messages in a conversation thread.

If the array is empty, tell the user there's nothing to process and stop.

2. Group by sender/company and auto-skip duplicates

Sort all fetched emails chronologically (oldest first). Then walk through them and assign each one a dedup key identifying the recruiter/company behind it:

  • Prefer the company name — pull it from the subject line, the body, or the sender's email domain (e.g. [email protected] → Ivo, a LinkedIn InMail pitching "Suno" → Suno).
  • If no company can be confidently identified, fall back to the sender's from_name (the human recruiter), since the same person may email from more than one address (e.g. a direct company address and [email protected]).
  • If neither is identifiable, fall back to from_email.

For each email, check whether its dedup key has already been seen earlier in this same run (including plain thread replies — Re: subject or quoted prior messages — which always share a key with the message they reply to).

  • The first (earliest) email for a given dedup key is a "primary" email — it goes through the normal interactive loop in step 4.
  • Every later email sharing that key is a duplicate or "bump" (a follow-up nudge, a repeat pitch for a different role at the same company, or the same recruiter reaching out from a different address). Auto-skip it without prompting: archive and mark read using the same commands as the Skip action in step 5, and print one line per auto-skipped email, e.g.:
Auto-skipped (duplicate of "<subject of primary email>"): <from_name> <<from_email>> — <subject>

Only primary emails proceed to step 3.

3. Load templates

Read all .md files in ~/.claude/skills/recruiter-reply/templates/. For each file, parse the YAML frontmatter to extract the name field. Sort the files alphabetically by filename. Build a numbered list with these templates followed by a fixed final option: Skip (no reply; leave email unread).

4. Interactive loop

For each primary email in chronological order:

a. Check whether this email is a follow-up in a longer thread. A follow-up is indicated by a Re: prefix in the subject, or by the email body containing quoted prior messages. If it is a follow-up, fetch the full thread:

~/.claude/skills/recruiter-reply/.venv/bin/python ~/.claude/skills/recruiter-reply/gmail_helper.py fetch-thread --thread-id "<thread_id>"

This returns a JSON array of all messages in the thread. Read the earlier messages and produce a 2–4 sentence Thread summary covering: the role/company being pitched, any compensation or remote/hybrid/in-office details mentioned, and how many times the recruiter has followed up.

b. Print the email as plain text in this format (no markdown blockquotes):

From: <from_name> <<from_email>>
Subject: <subject>

If a thread summary was produced in step (a), print it next:

Thread summary: <summary>

Then print the latest message body:

│ <body line 1>
│ <body line 2>
│ ...

c. Print the numbered option list built in step 3, followed by the prompt:

1. <template name>
2. <template name>
...
N. Skip

Enter a number:

d. Wait for the user to type a number in the chat and proceed accordingly.

5. Process each choice

For any template option (not Skip):

a. Read the selected template file. Strip the YAML frontmatter block (everything between the opening and closing --- lines) before using the body. b. Replace {name} with the sender's first name. c. If the template body contains {company}: replace it with the company name from the email. If you cannot determine the company name, ask the user before sending. d. Write the filled body to /tmp/recruiter_reply_body.txt e. Send the reply:

~/.claude/skills/recruiter-reply/.venv/bin/python ~/.claude/skills/recruiter-reply/gmail_helper.py send \
  --thread-id "<thread_id>" \
  --message-id "<gmail_message_id>" \
  --to "<from_email>" \
  --subject "<subject>" \
  --body-file /tmp/recruiter_reply_body.txt

f. Archive and mark read:

~/.claude/skills/recruiter-reply/.venv/bin/python ~/.claude/skills/recruiter-reply/gmail_helper.py archive --message-id "<gmail_message_id>"
~/.claude/skills/recruiter-reply/.venv/bin/python ~/.claude/skills/recruiter-reply/gmail_helper.py mark-read --message-id "<gmail_message_id>"

For Skip:

Archive and mark read (the email keeps its "recruitment" label, so it stays out of future runs):

~/.claude/skills/recruiter-reply/.venv/bin/python ~/.claude/skills/recruiter-reply/gmail_helper.py archive --message-id "<gmail_message_id>"
~/.claude/skills/recruiter-reply/.venv/bin/python ~/.claude/skills/recruiter-reply/gmail_helper.py mark-read --message-id "<gmail_message_id>"

What ships with it: 8 files

13.7 KB alongside SKILL.md, 1 of them executable

templates/

Gives 0 of the 12 instructions most hr recruiting skills give in ~1.3k tokens

Counted across 356 of the 357 authors here whose files we hold, read 2026-08-07

  • Quantify achievements with specific metricsin 14 of 356, across 6 files
  • Keep the resume under two pagesin 14 of 356, across 6 files
  • Request the full job description if not providedin 12 of 356, across 4 files
  • Extract keywords and prioritize job requirementsin 12 of 356, across 4 files
  • Stop and ask for clarification if required inputs are missingin 12 of 356, across 5 files
  • Map candidate experience to job requirementsin 11 of 356, across 3 files
  • Ask if the user wants adjustmentsin 11 of 356, across 3 files
  • Provide strengths and gap analysis after the resumein 10 of 356, across 2 files
  • Request candidate background details if not providedin 10 of 356, across 2 files
  • Format experience bullets as action verb plus resultin 10 of 356, across 2 files
  • Ask for missing inputs before startingin 10 of 356, across 9 files
  • Use exact job description terminologyin 9 of 356, across 1 file

Said here and by no other author read

  • Fetch unread emails with the recruitment label
  • Stop if no emails are found
  • Sort fetched emails chronologically
  • Assign each email a dedup key
  • Auto-skip duplicate emails without prompting
  • Load and parse all reply template files

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.