Find clear steps for a job that helps your business grow. Pick one guide, give your AI the facts, and check its work. These guides support the Content Factory, which turns your real stories into useful posts, pages, and ads.
Start below: search for a job, open its guide, and choose Copy first-run prompt. Paste it into your AI app and add the files and facts the guide asks for. Try one draft before you set up repeats. See the setup guide for app support and access.
The live library below shows current counts. Reviewed steps, reported completion, and proof from a real run mean different things. See which checks each task still needs.
Download the guides and first steps
Download all task guides (.zip). Unzip the file and open START-HERE.md. A ZIP is a folder of instructions; downloading it does not connect accounts or start work.
For a first try, use one guide and supply its requested inputs. Connect only the accounts that job needs. After one run works, set a schedule only if you want repeats, then check that the first timed run happened. For an app package, follow the install guide.
These guides are part of the Content Factory: each guide helps with one step, from gathering a story to posting it and checking the result. The Asset Tracker connects related work and resources.
These older notes describe past work. They do not certify the current library. Read the October 2 setup repair and its checks.
Importance bars, factory chain, and single-engine routing
Every task now carries a 1-5 importance score (max of frequency, ads/revenue, and gating) so a pixel install can outrank a long SOP. The dashboard shows the Content Factory line — Produce, Process in Descript, Post, Promote — with Digital Plumbing as Gate. Model routing is first-class for Claude-only and Grok-only operators; multi-engine is optional throughput on the same line. 123 incomplete skills inventoried; highest-value ads and access stations fleshed. How we closed the factory gap.
Monthly capability refresh — and the first entry this panel has been able to record since it was built
Reviewed the July 2026 releases across Anthropic, Google and OpenAI and rewrote the operating layer against them. Opus 5's per-turn reasoning-effort dial (July 24) became a routing rung of its own — raise effort before raising model — and every scheduled job now names a fallback tier, after Fable's two-and-a-half-week suspension in June proved a single-engine routing table is a single point of failure. Sonnet 5 replaces the smaller models as the default for any step that chains tool calls. The same run found three of three download buttons advertising hand-typed counts frozen months earlier: /dealcon offered 10 skills for a 19-skill pack, and both Task Library surfaces said 239 for a 247-file download. Counts are now derived from each pack's own VERSION.txt on every daily run. What broke and what we changed.
Gap in the record. One monthly cycle is missing from this record. The panel used to live inside the dashboard app itself, which moved to a host we hold no write credential for — so from June 10 to August 1 there was no way to append to it, while the daily "updated" badge above it kept stamping the correct date. The gap is left visible on purpose. A proof-of-recursion panel that hides the months it could not record is not proof of anything.
Task Library dashboard published
239 documented tasks across 13 categories, each a runnable skill.md, with the searchable dashboard and the downloadable pack. How the library was built.
Every monthly refresh appends an entry here, and this panel lives in this page's own content — not inside the dashboard app — so the job that is supposed to write to it always can. The changelog is the proof of recursion; a panel nothing can append to proves nothing.
Monthly capability refresh — and the first entry this panel has been able to record since it was built
Reviewed the July 2026 releases across Anthropic, Google and OpenAI and rewrote the operating layer against them. Opus 5's per-turn reasoning-effort dial (July 24) became a routing rung of its own — raise effort before raising model — and every scheduled job now names a fallback tier, after Fable's two-and-a-half-week suspension in June proved a single-engine routing table is a single point of failure. Sonnet 5 replaces the smaller models as the default for any step that chains tool calls. The same run found three of three download buttons advertising hand-typed counts frozen months earlier: /dealcon offered 10 skills for a 19-skill pack, and the Task Library surfaces mixed a 239-file hub count with larger versioned downloads. This page now labels the live registry, runnable ZIP, and archived bundle separately; upstream PR #1 makes the Task Library counts build-derived after merge. What broke and what we changed.
Gap in the record. One monthly cycle is missing from this record. The panel used to live inside the dashboard app itself, which moved to a host we hold no write credential for — so from June 10 to August 1 there was no way to append to it, while the daily "updated" badge above it kept stamping the correct date. The gap is left visible on purpose. A proof-of-recursion panel that hides the months it could not record is not proof of anything.
Task Library dashboard published
Legacy migration imported 239 local task-skill files across 13 categories. That count did not prove definitive-article coverage, ownership, or production reliability; current status and evidence fields expose those gaps. How the library was built.
Every monthly refresh appends an entry here, and this panel lives in this page's own content — not inside the dashboard app — so the job that is supposed to write to it always can. The changelog is the proof of recursion; a panel nothing can append to proves nothing.
Where this fits
Context is what the worker knows. A skill is the method it follows. A pack is a folder of methods. None of them does any work alone. Work happens when an AI worker follows a guide with the facts and access it needs, checks the output, and saves the next step. Set a repeat schedule only after a first run works.
- Context — the verified facts, goals, evidence, decisions, and working state kept in user-owned Markdown, Obsidian, Drive, or Git so any authorized model can pick up the same work.
- SkillYOU ARE HERE — one task written as steps, with required inputs, result checks, and a next handoff. Read its evidence before trusting it.
- Pack — a group of guides. App support and loading steps vary; follow the install guide.
- Agent — an AI worker with a defined job, tools, and permission to use them.
- Job — one attempt to get a result. A repeating job also needs a schedule, a place to run, and a check that it actually ran.
- Proof — a record of what happened, including failed or blocked work. Public release needs permission. Review the lesson before changing the guide.
The map: Content Factory · every asset: Asset Tracker · next door: How we build agents.