agentleFS
gabchessgabchess/gtm-campaign-launcher/skills/campaign-launcher/SKILL.md

campaign-launcher

Skill0 starsChanged 2 months agoMITView on GitHub

Run a cold-outbound campaign (a contact list, copy variants, and a multi-step sequence) through a QA gate before generating sending-tool configs. Blocks bad contacts, unresolvable merge tags, spam-trigger phrases, and a missing reply-stop condition, with a named reason for each. Use when the user has a campaign list CSV, email copy, and a sequence and wants it checked and turned into an Instantly or HeyReach config draft before upload. Trigger phrases -- "QA this campaign", "check my outbound list before I launch", "generate my sending config", "run preflight on this sequence".

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. lucaspaesbrazil-lab/agencia-digital-showcase★ 0same 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.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run the QA gate first, always:

    line 32
  2. Very specific:

    Every generated file is labeled "generated, review before upload" in its own content.

    line 75
  3. Distinctive:

    When you report the result of a run, use the gate's own PASS/BLOCKED language and named reasons.

    line 81
  4. Distinctive:

    Call any sending-tool API, upload a list, or send a message.

    line 91

≈ “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

Campaign Launcher

You take a campaign directory (a contact list, copy variants, and a sequence definition), run it through a QA gate, and only then generate draft configs for a sending tool. You do not invent contact data, and you do not generate a config for a campaign that failed the gate.

Inputs

A campaign directory with three parts. See fixtures/campaign/ for the shape this repo ships with.

  1. list.csv -- one row per contact: email, first_name, last_name, company, title, domain.
  2. copy/*.md -- one file per sequence step (or per A/B variant of a step), each a Subject: line plus a body using {{merge_tag}} slots that must match a list.csv column name.
  3. sequence.json -- ordered steps, each with a delay_days, an optional condition, and one or more variants (a file plus a split percentage for A/B tests). The sequence must declare "stop_on_reply": true at the top level.

Step 1: Never generate before the gate passes

Run the QA gate first, always:

python3 scripts/preflight.py fixtures/campaign

This checks three things, and reports every failure with a named reason instead of a generic "invalid":

  • List QA -- email syntax, the email's domain must match the domain column, no duplicate contacts, no missing required field. A bad row is excluded from the generated config, named by reason; it does not by itself stop the rest of the campaign from generating.
  • Copy QA -- every {{merge_tag}} in every copy file must be a real list.csv column, no phrase from the spam-trigger list, and the body must be under that step's word cap (cold-open 120, followup 80, breakup 60). Any hit here blocks the whole campaign -- a copy problem affects every contact, not just one row.
  • Sequence QA -- steps numbered in order starting at 1, sane delay_days (0 on step 1, 1-14 after that), a reply-stop condition present, and each step's A/B split summing to 100. Any hit here also blocks the whole campaign.

A campaign with zero copy/sequence problems still emits, even if one or more individual contacts were excluded for a bad row. A campaign with any copy or sequence problem emits nothing until that problem is fixed.

Step 2: Generate configs only on a clean gate

python3 scripts/emit_config.py fixtures/campaign

This re-runs the gate internally. On BLOCK, it prints the same report and exits 1 -- it writes nothing. On PASS, it writes three files to outputs/ (gitignored, never committed):

  • instantly-campaign.json -- shape modeled on Instantly's public campaign API docs.
  • heyreach-campaign.json -- shape modeled on HeyReach's public API docs.
  • launch-checklist.md -- a human checklist to run before uploading either file anywhere.

Every generated file is labeled "generated, review before upload" in its own content. Neither script makes a network call. Nothing in this repo uploads, sends, or signs into any sending tool.

Step 3: Report the gate's own words

When you report the result of a run, use the gate's own PASS/BLOCKED language and named reasons. Do not paraphrase "blocked" into "looks mostly fine" or hand-wave a failure as "should be okay to ship."

What this skill will not do

  • Generate a config for a campaign with an unresolved copy or sequence problem.
  • Invent a merge tag, a spam-safe rewrite, or a reply-stop condition the user did not provide.
  • Call any sending-tool API, upload a list, or send a message. Output is a reviewable draft on disk.

Related

  • scripts/preflight.py -- the QA gate described above.
  • scripts/emit_config.py -- the config generator, gated by preflight.
  • scripts/self_check.py -- validates this repo's own fixtures and structure are intact, including that the good fixture passes the gate and the broken fixture does not; run it after editing fixtures/.
  • fixtures/campaign-broken/ -- the NO-GO case: a duplicate contact, an unresolvable merge tag, and a missing reply-stop condition, for reference.

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.