Happy Path Works

The first run is the finished run.

Workflow systems for Claude — skills, setup guides, and worked examples, each measured against a quality floor published on this site. Read the documentation before you buy anything.

Get the free skill audit Document Ops — November 10

skill_audit — skill-audit SKILL.md scripts: skill_audit.py [D] Upload-safe description PASS description is 698 of 1024 chars and holds no XML-tag shape. [R] References resolve PASS All 1 bundled path(s) resolve. (1 workspace path(s) skipped — not this skill's to resolve) [1] Names the gap PASS A gap/failure-mode statement is present. [2] Deterministic entry PASS A deterministic entry (prefix or unambiguous condition) is present (`audit:`). [3] Enforced hard-fails PASS Hard-fails are present and the skill ships a script (skill_audit.py). [4] Verify before voice PASS A hard constraint is present and a verification step is described. [5] Loud failure, named exit PASS Scope boundaries name an explicit exit (if X, refuse and do Y). [6] No prior-chat references PASS No phrasing that leans on an earlier chat ("as we discussed", "like last time"). [7] Deciding example PASS A worked example is present and shows a catch, not just the happy path. FAIL 0 REVIEW 0 NOT CHECKABLE 0 PASS 9 VERDICT: PASS Every gate came back clean.Not graded: whether the skill works in a fresh session. The lint reads files; only a run shows what a model does with them. -> Open a session where your own project instructions do not load, install the skill from its archive, give it a real task that starts with `audit:`, and check what it does against what the skill promises.

The free audit tool, run against its own skill file. Every gate it can check came back clean. The last line of the run is the part to read: a lint reads files, so it cannot tell you whether a skill works in a fresh session. The tool prints that limit instead of leaving it out, and so does this page.

What you get

Three things, not one.

Skills

Files Claude loads and runs. Deterministic triggers, hard-fails backed by real checks, named exits when a boundary is hit. Not prompts you paste.

Setup guide

Install, configure, verify — written per skill, not one README for the bundle. Published here, readable before purchase.

Worked examples

Each one shows the skill catching a case that would otherwise go wrong. An example of the obvious case succeeding is decoration.

The floor

Seven gates, and a check that returns an exit code.

The floor is published in full, including the anti-patterns that flatter an audit and help nobody. Four of the seven, starting with the one that costs the most to clear:

Enforced hard-fails

Every stated hard-fail is backed by a check that returns a signal before the skill emits. One unenforced hard-fail fails the gate.

exit 1 # FAIL — 3 BLOCK hits, nothing written

Verify before voice

A countable constraint is measured first. The claim is written around the number, not around an estimate of it.

count = len(desc) # then write the sentence around count

Loud failure, named exit

Boundaries say what to do instead, not "try to stay in scope".

if no brand config: refuse, print setup step 2

Deterministic entry

A hard trigger fires the skill. A keyword list is a supplement, never the mechanism.

trigger: any message beginning with "audit:"

Read the full floor →

Systems

Available now

Skill audit

Point it at a skill file. It grades against the seven gates, names what is missing on each failure, and exits 0, 2, 1, or 3. Free, source public, no email required to run it.

1 skill · free

Launches November 10, 2026

Document Ops

Five skills that turn a process you already run into a document someone else can follow — a branded SOP, a one-page job aid, a meeting agenda that eats its own minutes, a decision record built for sign-off. Your brand configured once, read by all of them.

5 skills · $79

Launches January 12, 2027

Session Discipline

What keeps long work coherent across threads: handoff blocks that survive a cold restart, a capture system that turns a passing thought into a scheduled task, a daily brief that assembles itself, and an audit that goes deeper than the free one.

4 skills · $49

All systems and what's planned →

The config layer

Set it once. Every skill reads it.

Three run-once intake skills hold the things every other skill needs and none of them should ask you twice — your visual identity, your register, your business facts. A skill that needs your logo reads the config. A skill that needs your terms reads the config. Each pack you own makes the next one cheaper to set up and worth more to run.

Brand-kit intake — visual identity: logo, colors, fonts, header style. Ships in Document Ops

Voice-capture intake — register and phrasing. Planned

Business-profile intake — company facts, terms, numbering schemes. Planned

The documentation is not behind the purchase.

Setup guides and worked examples for everything sold here are readable without an email address. If the docs do not convince you the system is built, do not buy the system.

Read the documentation →