caffeinity-routing
Picks which caffeinity skills to run, in what order, for a goal that spans more than one of them, and says what each step hands the next: build and publish a skill, get a project found, pass a release gate, grow organic search traffic, run a job search, make a decision, learn a skill, run cold outbound, launch a one-person product, set up a coding agent for a repo, or set up and keep the skills themselves healthy. Checks which of those skills are installed and names the missing ones instead of improvising their work. Use whenever someone asks which skill to use, what to run first, how the skills fit together, or describes a multi-step outcome such as "take this idea to a published skill" or "get my site ready to ship and ranking", even if they never name a skill. Not for a request one skill already owns (load that skill directly), and not for doing a step's work itself: the named skill does it.
- License 许可
- MIT
- 最近更新
- 2026-11-14
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- caffeinity-routing
- version
- 1.1.0
- description
- Picks which caffeinity skills to run, in what order, for a goal that spans more than one of them, and says what each step hands the next: build and publish a skill, get a project found, pass a release gate, grow organic search traffic, run a job search, make a decision, learn a skill, run cold outbound, launch a one-person product, set up a coding agent for a repo, or set up and keep the skills themselves healthy. Checks which of those skills are installed and names the missing ones instead of improvising their work. Use whenever someone asks which skill to use, what to run first, how the skills fit together, or describes a multi-step outcome such as "take this idea to a published skill" or "get my site ready to ship and ranking", even if they never name a skill. Not for a request one skill already owns (load that skill directly), and not for doing a step's work itself: the named skill does it.
02 资源文件 3
- 参考文档references/trigger-cases.md2.8 KB
- 参考文档references/workflows.md12.8 KB
- 脚本scripts/route_check.py5.6 KB
03 SKILL.md 正文
caffeinity-routing
A goal like "ship my site" or "get hired" crosses several skills, and each one only knows its own
edges. This skill holds the map: twelve numbered brews (Brew 1 to Brew 12; older docs call them W1 to W12), each an ordered list of
skills with what every step receives and what it hands on. Other skills cite these numbers
(release-gate-chair implements Brew 3 · Espresso; seo-crawl is the first step of Brew 4 · Cold brew), so keep them stable.
Routing is a decision, not the work. Name the workflow, the first step and its input, then hand off. Doing a step's job here produces a worse version of a skill that already exists.
Ground rules
- One skill owns a single-step request. If the request fits one skill's description, route straight to it and stop. A workflow is for goals that need two or more.
- Name the brew by number and name, and the steps by skill name. "Brew 4 · Cold brew, step 1: seo-crawl on https://example.com" is checkable; "do some SEO first" is not.
- Never fill a missing step yourself. Run
scripts/route_check.pyand report what is not installed with its install command. A step skipped for lack of a skill is reported as skipped, not as passed. - Carry outputs forward verbatim. Each step's output is the next step's input, as listed in
references/workflows.md. Summarizing between steps loses the evidence the next skill checks. - Respect each skill's own gates. A release gate that says HOLD, a keyword map that is not signed off, or a draft the user has not confirmed stops the workflow there. Say so; do not route around it.
Workflow
-
Classify the goal. Read the request against the workflow table below. If it fits one skill, route there and stop. If it fits more than one workflow, pick the one whose first step the user can start now and say which you set aside.
-
Check what is installed. Run:
python scripts/route_check.py espresso # one brew, by name, number or old W id python scripts/route_check.py --all # every brewIt reads
references/workflows.mdand looks for each skill'sSKILL.mdunder the usual skill folders (~/.claude/skills,~/.qoder/skills,~/.cursor/skills,~/.agents/skills,.claude/skillsand plugin caches). Add--path <dir>for any other location. -
Lay out the route. Open
references/workflows.mdfor the chosen workflow and write one line per step: skill, what it receives, what it hands on, and whether it is installed. Mark the chain skill if one already packages the steps (Brew 1 · Pour-over hasskill-launch, Brew 3 · Espresso has therelease-gate-*forks, Brew 4 · Cold brew hasseo-polish,keyword-resolverandseo-sweep, Brew 10 · Descale hastune-and-shipfor one published skill at a time). -
Hand off step 1. Load the first skill with its input. When a step finishes, come back to the route, record the step's verdict or output location, and hand off the next one.
-
Close the loop. When the last step finishes, list each step as done, skipped (with the reason) or stopped at a gate, plus the one next action.
The brews
| # | Goal | Steps, in order |
|---|---|---|
| Brew 1 · Pour-over | Build a skill and make it installable | draft-skill or prompts-to-skill → tune-skill → prompts-to-skill-release → skill-to-mod (only if it must run at a fixed moment) |
| Brew 2 · Bloom | Get a repo, skill or plugin found | findable → publish-pipeline → upstream-check → get-indexed (if new pages went live) |
| Brew 3 · Espresso | Decide whether a site is safe to launch | web-preflight → pre-ship-security-red-team → pre-ship-performance-audit → release-gate-chair (or its -ledger and -validator forks) |
| Brew 4 · Cold brew | Grow organic search traffic | seo-crawl → keyword-guard → wordsmith → pre-ship-performance-audit, platform-content, get-indexed (each if it applies) |
| Brew 5 · French press | Land a job | recruiter-lens → applicant-kit → application-pilot → career-track (for an offer, promotion or pivot) |
| Brew 6 · Cupping | Make a decision | research-brief → data-readout (if there is data) → crash-test or do-better → strategy-memo |
| Brew 7 · Slow drip | Learn a skill | topic-primer → study-system → course-builder (only when teaching a group) |
| Brew 8 · Nitro | Run cold outbound | cold-email-list-gate → cold-email-ops → sales-copy |
| Brew 9 · Moka pot | Launch a one-person product | solo-income-blueprint → site-prompt → production-code → paper-trail → Brew 3 · Espresso → platform-content |
| Brew 10 · Descale | Maintain the skill family | tune-skill (regress.py) → prompts-to-skill-release (if behaviour changed) → upstream-check |
| Brew 11 · Estate roast | Stand up or audit a skill estate | skill-estate (eight gated phases over the whole collection) |
| Brew 12 · Dial-in | Set up a coding agent for a repo | portable-harness or harness-audit → setup-claude-advisor (if quota is the constraint) → diagnose-usage (if usage already drained) → choose-effort (before a large task) |
Per-step inputs, outputs and stop conditions are in references/workflows.md. Open it at step 3,
and whenever a user asks what one skill needs from the previous one.
references/trigger-cases.md holds the hand-run trigger cases; open it when tuning this
description or adding a workflow.
Report
Lead with the route: workflow number, then one line per step with its status. Then the missing skills with install commands, any gate that stopped the run, and the one next action. Say which steps you only planned and which actually ran.
04 同仓库更多 Skills
tune-and-ship
Takes one published skill at a time from as-is to tuned, fact-checked, released and live, then moves to the next one in the catalog. Each run tunes the skill with tune-skill, checks every version number, flag, price and command it states against the primary source, re-reads its siblings' real SKILL.md files, polishes the listing with findable-polish, writes the version and release notes, opens one PR, and after the merge confirms the site page and marketplace copy are live. Use whenever the user wants to fine-tune their other skills, work through the catalog, give a published skill the full treatment, fact-check a skill's claims, or ship a tune instead of drafting it, even if they never say "ship". Not for a tune with no release (use tune-skill), drafts with nothing pushed (use tune-and-polish), a new skill (use skill-launch), regressions from the nightly health check (use caffeinity-routing's Descale brew), or a whole collection (use skill-estate).
skill-estate
Takes a whole collection of agent skill repos from scattered to published, consistent and self-maintaining, in gated phases: inventory every repo, reconcile site, marketplace and README listings, test and tune the family, close workflow gaps, wire publishing so new skill repos go live on their own, bundle a plugin with agents, add trigger tests and directory listings, and leave a maintenance loop and a map. Each phase hands off to the skill that owns it and ends in PRs. Use when someone's skills site and GitHub disagree, wants all their skill repos audited or cleaned up, wants new skills to publish themselves, or wants that cleanup repeated on another org, even if they never say estate. Not for one skill (use tune-skill), one listing (use findable), the site pipeline alone (use publish-pipeline), or weekly upkeep (use caffeinity-routing's Descale brew).
paper-trail
Drafts the policy pages a small product needs before it takes money or data (privacy policy, terms of service, refund policy, cookie notice) from a fact sheet about what the product actually collects, stores, shares and charges, then checks the drafts for placeholders, missing sections and facts nobody supplied. Use when someone needs a privacy policy, terms, refund or cookie page for a site, app, SaaS, newsletter or digital product, asks what legal pages they need before launch, or wants existing policy pages checked against what the product really does, even if they never say "legal". Not legal advice and not for contracts, NDAs or employment documents (use a lawyer), and not for the launch plan itself (use solo-income-blueprint) or the site build (use site-prompt).
course-builder
Designs a course for a group of learners (a class, cohort, workshop series, bootcamp or internal training): measurable outcomes, a module sequence with prerequisites, activities and assessments aligned to each outcome, a time budget that fits the hours available, and a rubric, then checks the plan with a script so no outcome goes untaught or unassessed. Use when someone is teaching others and asks for a curriculum, syllabus, course outline, module plan, cohort or workshop structure, or wants an existing course checked for gaps and overload, even if they never say "curriculum". Not for one person's own learning plan (use study-system), for being taught the topic (use topic-primer), or for marketing the course (use sales-copy).
wordsmith
Write, draft, rewrite, edit, proofread, polish and improve text while preserving the author's intended meaning, facts and voice. Use when the user wants clearer, shorter or better-toned writing; grammar, structure or flow fixes; emails, website or marketing copy, headlines, outlines or summaries; a different reading level; an accessibility, inclusivity or accuracy review; or content prepared for publication, even if they never say "edit". Also handles /wordsmith commands such as draft, critique, audit, tighten, tone and final. Not for resumes, CVs or LinkedIn profiles (use recruiter-lens if installed), not for code or code comments, and not for building .docx or .pdf files (write the text here, then use the file skill).
web-preflight
Use when shipping, deploying, publishing, launching, reviewing, or pre-flighting a website or API and you want a lightweight check for accidental exposure of secrets, source or configuration files, admin and debug surfaces, unsafe redirects, missing security controls, and misleading 200 responses. Trigger phrases include "before we ship", "security check", "preflight", "audit the site", "check the production URL", "is my .env exposed", and "why does /.env return 200", including when access-log or analytics evidence is supplied. Checks only sites the user owns or is authorized to assess, with bounded GET and HEAD requests and a fixed compact path set; distinguishes genuine resources from SPA catch-all and blanket WAF responses. Not for penetration testing, exploitation, credential testing, broad port scanning, or application-code review; use a dedicated authorized security-testing or code-review workflow instead.