a11y-check
Accessibility audit, scoped to the surfaces a product actually has. Detects web / rendered_markdown / terminal / native_app / video_audio / document /…
Render a completed diamond's work as a shareable, factual one-page "work receipt" (problem framed, assumptions tested, what was killed vs kept) carrying an origin line + a volitional onward-handoff data flow. Read-only. Consults the attribution registry per
$ npx -y skills add haabe/mycelium --skill receipt-render --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/receipt-renderContext preview
The summary Claude sees to decide when to auto-load this skill.
Render a completed diamond's work as a shareable, factual one-page "work receipt" (problem framed, assumptions tested, what was killed vs kept) carrying an origin line + a volitional onward-handoff data flow. Read-only. Consults the attribution registry per
name: receipt-render
description: Render a completed diamond's work as a shareable, factual one-page "work receipt" (problem framed, assumptions tested, what was killed vs kept) carrying an origin line + a volitional onward-handoff data flow. Read-only. Consults the attribution registry per `${CLAUDE_PLUGIN_ROOT}/engine/render-conventions.md` consent + privacy gate. Follows the render fleet's shared conventions + consent gate (Check 43 applies via the `-render` name); a standalone shareable-artifact generator, NOT a `/render` dispatcher target. Audience fixed `external` (a hand-it-over artifact).
metadata:
instruction_budget: "65"
framework_dependency: "mycelium"
framework_dependency_note: "Reads .claude/diamonds/active.yml + .claude/canvas/{archived-solutions,opportunities,cycle-history}.yml + .claude/harness/decision-log.md + attribution registry (env var $MYCELIUM_ATTRIBUTION_REGISTRY preferred; .claude/memory/attribution-registry.yml fallback). Standalone use will fail without a completed diamond present."
identifier_exposure: "YES"Read-only render of a **completed diamond** as a shareable, factual one-page **work receipt** — the problem as it was framed, the assumptions tested, what was killed and what was kept. It follows the render fleet's shared conventions and consent gate (`${CLAUDE_PLUGIN_ROOT}/engine/render-conventions.md`; Check 43 applies via the `-render` name), but unlike the diagram specialists (`diamond-render` / `ost-render` / `cycle-render`) it produces a shareable prose artifact and is **not** a `/mycelium:render` dispatcher target.
The receipt is a **carrier the user may choose to hand to a colleague**. It is generated fully **locally** — nothing is transmitted. Its footer carries a volitional data flow (an origin line + a pre-filled GitHub Discussion deep-link) so that, *if the user chooses*, the work can surface back to the maintainer. The skill never posts anything; a human reviews a draft and submits it.
A silent adopter is invisible to a solo maintainer. This skill does not generate adoption and does not phone home. It gives a real piece of completed work a shareable shape with a low-friction, non-leading way for the user (or a colleague they hand it to) to surface themselves — if they want to. It surfaces **behaviour** (what was built), never solicits **praise** (whether it was good).
**Declared**: YES
| Canvas file | Identifier-bearing fields | Frequency | |---|---|---| | `.claude/canvas/archived-solutions.yml` | `reason` / `notes` prose (may name a tester whose feedback killed a solution) | low | | `.claude/canvas/opportunities.yml` | `evidence_sources`, `notes` prose | low | | `.claude/harness/decision-log.md` | decision prose filtered to this diamond (may name contributors) | low–medium | | `$MYCELIUM_ATTRIBUTION_REGISTRY` env var (canonical) or `.claude/memory/attribution-registry.yml` (fallback) | `people:` block with `name`+`consent`+`note` | read-only consultation; never rendered |
Per `${CLAUDE_PLUGIN_ROOT}/engine/render-conventions.md#hard-rule-consent--privacy-gate`. The registry lives in roadmap-private memory by design — it must never ship in a generated receipt.
The receipt is a **hand-it-over artifact**, so its audience is fixed `external` (stricter than the render fleet's `cohort` default): only `public_ok` renders literal; everything else redacts. A receipt that leaks a tester's name or a private project name to a colleague is a privacy incident, and the user may not notice before sharing — so the gate is conservative by construction, not by operator vigilance.
Per `engine/render-conventions.md#anon-label-convention`:
Numbering resets per render; emit an anon-mapping footnote when redaction occurred.
| Value | Render behavior | |---|---| | `public_ok` | Render literal. Append carve-out footnote pointer if entry has non-empty `note:`. | | `generic_only` | Redact to anon-label (identity is roadmap-private; this artifact is external). | | `unknown` | Treat as `generic_only` (conservative). | | (not in registry) | Fail loud unless `--no-identifiers=true`. |
**Project names** are not in the registry; a receipt MUST NOT name a third party's private project. If the source work references a private project name, redact to a generic descriptor and surface a warning. (See the Frida carve-out precedent in `engine/render-conventions.md#carve-out-notes`.)
**public_ok → literal**: decision-log names `Drew`, registry `{name: "Drew", consent: public_ok}` → renders `Drew` + carve-out footnote pointer.
**generic_only → anon-label**: `archived-solutions.yml#reason` names `Daniel`, registry `{name: "Daniel", consent: generic_only}` → renders `peer-practitioner-1` + anon-mapping footnote.
**not in registry → fail loud**: a receipt source names `Random Name` with no registry entry → fail-loud with three fix options (add to registry, re-run with `--no-identifiers`, edit the source canvas).
A harness that asks who this is for before the agent writes code. Built on Claude Code, where the gates are structural. The files and skills port to opencode, Codex and Cursor. Outcome over output. You know how this goes.
Accessibility audit, scoped to the surfaces a product actually has. Detects web / rendered_markdown / terminal / native_app / video_audio / document /…
Bring Mycelium into a project that already has code. Detects that the repo predates the framework, asks before touching anything, then reads the codebase to…
Design the smallest viable test to validate or invalidate a critical assumption. Based on Torres's assumption testing framework, organized by Gilad's AFTER…
Use before any research activity or significant decision. Reviews cognitive biases relevant to the current stage.
Use to evaluate whether current work aligns with Better Value Sooner Safer Happier. Run at diamond completion and periodically.
Lint canvas files for staleness, missing fields, inconsistent evidence types, and orphaned references. Run periodically or before major transitions.