profession@di-atomic/propagate · v0.1.0 · internal

Launch Engineer

propagate@profession.md

Runs one bounded launch end to end, from any IDE. It acquires its own toolkit and proves it works before it starts, scores whether the release earns a launch at all, then seeds one card that every surface reads from. The hero renders exactly once.

Chalkboard sketch: a figure seen from behind raises a hand in a stop gesture toward a row of four squares, one filled gold. Hand-lettered label: LAUNCH ENGINEER.

This one is internal tooling — the role I equip when Di-Atomic, SpiderIQ or OPVS launches its own work. If you want the same craft pointed at your releases, the client-facing sibling is delivery-manager.

Built by Di-Atomic Marketing & compliance agency Dogfooded on our own launches 7-language team Security review 100/100, no findings
16linked skills — 7 CALL, 9 READ
5workflows, bootstrap first
16definition-of-done checklists
23explicit out-of-scope lines

What this role owns

One launch, start to finish — and the judgment call that comes before it. Not every release earns a launch.

Bootstrap before you workAn IDE session starts with an empty toolkit. Nine skills get extracted with their layer files intact; seven MCP surfaces get probed read-only. A half-equipped launch fails at publish time, which is the most expensive place to find out.
Score before you produceVerify the real delta, run the newsworthiness rubric, record score, tier, fired signals and any override with its reason. A release scoring 0–2 gets a changelog line and an honest "this isn't news."
One card, one heroThe launch card is the shared state. Every surface reads it, adds its piece, passes it on. The hero renders once and every consumer reuses it by URL — a second render of the same asset is a defect.
Verify live, then say doneEvery published surface gets fetched and grepped for its own marker before it counts as shipped. A 200 is not proof. Production deploys stay a human call.

toolkit-bootstrap

The intake. Gate on the CLI floor, extract the nine READs with their references and scripts, wire and probe the seven CALLs. Names the gap loudly if one can't be closed.

launch-intake

The gate. Resolve which brand you're shipping for, verify the delta is real, score it, then commit to a surface list or decline. Declining is the most valuable thing this role does.

seed-launch

The seeder. Create the card, seed the release facts, lay down every stage upfront, produce the blog body and render the hero exactly once, publish, hand off.

work-surface

The worker. One surface only. Read the card first, self-gate on inputs you didn't produce, reuse hero and voice and hook as given, merge-write only your own keys.

stalled-launch

The escalation. Stop cleanly with a named gap. A launch that halts is recoverable in minutes; one that improvises around a missing input ships something wrong nobody catches until a customer does.

three recipes

skill-launch, profession-launch, product-release — each fixes the surface list and the house rules for what is being launched.

The toolkit — 16 skills, linked not copied

A profession doesn't bundle its skills. It references them by name and version, and they stay independently owned and separately versioned. The split below is the one thing worth learning: READ skills are the method, CALL skills are the capability. Installing a SKILL.md gives you documentation. Ability comes from the MCP server plus the brand's own grants.

Chalkboard, two labelled groups. Under READ, a loose scatter of open books connected to nothing. Under CALL, a row of plugs each wired by cable to one shared gold socket rail.
READ — 9 guidance skills, executed by your own model
CALL — 7 capability skills, reached over MCP
@opvs-ai/agentboardThe launch card itself. Built by OPVS.
@opvs-ai/agentdocsSurface bodies, versioned. Built by OPVS.
@opvs-ai/agentmemoryWhich hooks converted last time. Built by OPVS.
@opvs-ai/opvs-protocolThe evergreen handoff. Built by OPVS.
@opvs-ai/professionThe guild hall. Built by OPVS.
@spideriq/publish-skillsPublishes live to the resolved domain. Built by SpiderIQ.
@spideriq/gateway-skillsThe render engine. Built by SpiderIQ.

Equip it

equipopvs profession equip @di-atomic/propagate

Your first thirty days: run toolkit-bootstrap before anything else and let it fail loudly if a CALL surface isn't wired. Then let the gate decline something. The first time it tells you a release doesn't earn a launch and you agree with it, the role has already paid for itself.

Two things it will never do. It won't fire a production deploy — it publishes surfaces, deploys to preview, and leaves the site-wide push to a human. And it won't touch client work.

Di-Atomic

We build the launch machinery we sell

Every @di-atomic skill and profession ships through this exact pipeline, on our own site, before it goes anywhere near a client. If you want a launch process that produces a record instead of five browser tabs, that's the conversation.

Book a 30-min call with Di-Atomic

Launching for clients rather than yourself? See delivery-manager.