Ost render

OST Render

Read-only render of .claude/canvas/opportunities.yml as a diagram or structured list. Second specialist in the render fleet (after /mycelium:diamond-render). See ${CLAUDE_PLUGIN_ROOT}/engine/render-conventions.md for shared conventions.

When NOT to use

• To BUILD or update the OST from research data → /mycelium:ost-builder. This skill is read-only.
• For cross-cutting opportunity→solution→cycle traceability view → dispatcher's /mycelium:render --view traceability (deferred to Phase 4a–4d research-first methodology per architecture draft §10.2).
• For ICE scoring of solutions → /mycelium:ice-score.

Identifier exposure

Declared: YES

Scope (canvas surfaces touched)

Canvas file	Identifier-bearing fields	Frequency
.claude/canvas/opportunities.yml	evidence_sources (URL-or-name strings); notes prose	low (~1 in 10 cites a named individual rather than a URL)
$MYCELIUM_ATTRIBUTION_REGISTRY env var (canonical) or .claude/memory/attribution-registry.yml (fallback)	people: block with name+consent+note per entry	read-only consultation; never rendered

Per ${CLAUDE_PLUGIN_ROOT}/engine/render-conventions.md#hard-rule-consent--privacy-gate. Registry lives in roadmap-private memory by design — upstream Mycelium repo is public; registry contents must not ship there.

Rationale

OST shape is user-need-shaped (abstract opportunities and solutions), so identifier exposure is incidental rather than intrinsic. Identifier classes that can appear: cohort testers, peer practitioners, named external sources. All three go through the consent gate; external sources are typically public_ok already, but registry consultation is non-skippable to avoid drift.

Anon-label convention

Per engine/render-conventions.md#anon-label-convention:

• Cohort tester → cohort-tester-N
• Peer practitioner → peer-practitioner-N
• Unknown identifier class → participant-N

Numbering resets per render; emit an in-render mapping footnote when redaction occurred so the operator can audit which anon label maps to which registry entry.

Consent value semantics

Per engine/render-conventions.md#consent-value-semantics:

Value	Render behavior
public_ok	Render literal. Append carve-out footnote pointer if entry has non-empty note:.
generic_only	Redact to anon-label. Queue anon-mapping footnote.
unknown	Treat as generic_only per registry README.
(not in registry)	Fail loud unless --no-identifiers=true.

Worked examples

public_ok → literal: evidence_sources: ["Drew Hoskins LinkedIn 2026-05-11"], registry entry {name: "Drew", consent: public_ok, note: "..."} → renders as Drew Hoskins LinkedIn 2026-05-11 + carve-out footnote pointer.

generic_only → anon-label: evidence_sources: ["Daniel Bentes install report"], registry entry {name: "Daniel", consent: generic_only, ...} → renders as peer-practitioner-1 install report + anon-mapping footnote.

Already-anon canvas label → preserve: evidence_sources: ["cohort-tester-3 session 2 transcript"], canvas already anon-labeled → renders as cohort-tester-3 session 2 transcript (preserve canvas-side anonymization; do not look up).

Identifier absent from registry → fail loud: evidence_sources: ["Random Name DM 2026-06-05"], no entry → the WHOLE render is blocked — emit NO artifact (not a partial mindmap with the offending entry redacted or omitted), and give the fail-loud message with three fix options (add to registry, re-run with --no-identifiers, edit canvas). Clarified v0.56.0 after a dogfood run read this line as per-entry and rendered around the unregistered source: an unresolved consent state means the render's exposure declaration cannot be made honestly, so nothing ships until the user resolves it. Naming the offending source in the owner-facing explanation is fine (project-private context); the artifact is what must not exist.

Fixture pointer

• tests/bash/fixtures/ost-render/redaction-public-ok-literal.yml
• tests/bash/fixtures/ost-render/redaction-generic-only-anon.yml
• tests/bash/fixtures/ost-render/redaction-unknown-treated-as-generic.yml
• tests/bash/fixtures/ost-render/redaction-no-registry-entry-fail-loud.yml
• tests/bash/fixtures/ost-render/redaction-carve-out-note-footnote.yml

Preflight: Read sources

1. Read .claude/canvas/opportunities.yml with the Read tool. Full read for emit; not limit:1.
2. Read the attribution registry per path-resolution order in engine/render-conventions.md#registry-path-resolution: $MYCELIUM_ATTRIBUTION_REGISTRY env var first; fall back to .claude/memory/attribution-registry.yml. Registry root key is people:; each entry has name, consent, optional note. If registry absent, surface ⚠ no attribution-registry — consent-redaction not enforceable; treat output as roadmap-internal warning in the render header.
3. Note the source's canvas-state timestamp per engine/render-conventions.md#canvas-state-timestamp-resolution: _meta.last_validated if present, else top-level last_updated:.

Arguments

Arg	Default	Values	Effect
--format	mermaid	mermaid | ascii | markdown-list | json	Output format. markdown-table is NOT supported (trees don't map to tables); fail loud per engine/render-conventions.md#format-support-negotiation-global-rule.
--shape	mindmap	mindmap | flowchart-td	Mermaid diagram shape. flowchart-td opt-in for users who prefer directed-graph rendering or whose target renderer doesn't honor mindmap palette.
--theme	base	base | dark	Theme. dark is the WCAG-by-construction opt-in per engine/render-conventions.md#wcag-aa-theme-convention.
--root-outcome	null	opportunity ID	Render sub-tree rooted at this opportunity. Fail loud if ID does not exist.
--include-status	all	active | archived | closed | resolved | all	Filter by lifecycle state.
--show-ice	true	bool	Suffix ICE score on solution nodes (drop if zero).
--show-confidence	false	bool	Suffix confidence on solution nodes.
--no-identifiers	false	bool	Force all name references to redact to anon-labels regardless of consent state.

Workflow

Step 1: Parse + filter

Read opportunities.yml. Build in-memory tree:

• Root = desired_outcome (top-level field).
• Branches = top-level opportunities.
• Leaves = solutions per opportunity.
• Apply --include-status filter (drop entries that don't match).
• If --root-outcome <id>, prune tree to subtree rooted at the given ID; fail loud if ID absent.

Step 2: Consent check on identifier-bearing fields

Per engine/render-conventions.md#hard-rule-consent--privacy-gate. For every evidence_sources entry AND every named-individual mention in notes prose:

• Skip URL-shaped entries (http://, https://).
• Skip file-path-shaped entries (*.yml#..., ../*, .claude/*).
• Skip already-anon canvas labels (cohort-tester-N, peer-practitioner-N, participant-N patterns).
• For name-shaped remaining entries: look up first-name token in the registry's people: block.
• Apply consent value semantics per the table above.
• Maintain consistent anon-label numbering across the render (same registry entry → same N).

Step 3: Apply suffixes + escape

• ICE suffix if --show-ice=true and ICE present.
• Confidence suffix if --show-confidence=true.
• Escape labels per engine/render-conventions.md#mermaid-label-escape-rules.

Step 4: Emit by format

Format mermaid (default) — Mermaid mindmap with WCAG AA theme.

Use frontmatter config syntax per engine/render-conventions.md#mermaid-frontmatter-syntax-preferred. Default uses Material Design palette (verified-working across mermaid.live, mermaidchart.com, Obsidian). --theme dark opt-in switches to Mermaid's built-in dark theme. --shape flowchart-td opt-in switches to directed-graph rendering for renderers that don't honor mindmap palette overrides.

---
config:
  theme: base
  themeVariables:
    cScale0: '#FDD835'
    cScaleLabel0: '#000000'
    cScale1: '#42A5F5'
    cScaleLabel1: '#FFFFFF'
    cScale2: '#66BB6A'
    cScaleLabel2: '#000000'
    cScale3: '#EF5350'
    cScaleLabel3: '#FFFFFF'
    cScale4: '#AB47BC'
    cScaleLabel4: '#FFFFFF'
    cScale5: '#26A69A'
    cScaleLabel5: '#000000'
    cScale6: '#FFA726'
    cScaleLabel6: '#000000'
    cScale7: '#8D6E63'
    cScaleLabel7: '#FFFFFF'
---
mindmap
  root((Desired outcome))
    opp-001[Opportunity 1]
      sol-001a[Solution 1a ICE 250]
      sol-001b[Solution 1b ICE 180]
    opp-002[Opportunity 2]
      sol-002a[Solution 2a ICE 90]

Critical: cScale0/cScaleLabel0 are wasted (off-by-one in mindmap source: section indexing starts at cScale1). Setting them is harmless belt-and-suspenders. cScale1 is the first rendered section's color.

Format ascii — terminal-friendly tree:

Desired outcome
├── opp-001: Opportunity 1
│   ├── sol-001a: Solution 1a (ICE 250)
│   └── sol-001b: Solution 1b (ICE 180)
└── opp-002: Opportunity 2
    └── sol-002a: Solution 2a (ICE 90)

Format markdown-list — hierarchical list:

- Desired outcome
  - opp-001: Opportunity 1
    - sol-001a: Solution 1a (ICE 250)
    - sol-001b: Solution 1b (ICE 180)
  - opp-002: Opportunity 2
    - sol-002a: Solution 2a (ICE 90)

Format json — external-system integration:

{
  "schema_version": 1,
  "render": "ost",
  "source": ".claude/canvas/opportunities.yml",
  "source_last_validated": "<YYYY-MM-DD>",
  "tree": {
    "outcome": "Desired outcome",
    "opportunities": [
      {
        "id": "opp-001",
        "title": "Opportunity 1",
        "status": "active",
        "solutions": [
          {"id": "sol-001a", "title": "Solution 1a", "ice": 250, "confidence": null}
        ]
      }
    ]
  },
  "redactions_applied": [],
  "dropped_fields": ["four_risks", "evidence_sources", "assumption_tests", "notes", "decision_log_refs"]
}

Step 4b: Validate the emitted Mermaid (MANDATORY for mermaid format)

Pipe the block you just emitted through the static validator before showing it:

printf '%s' "$DIAGRAM" | python3 ${CLAUDE_PLUGIN_ROOT}/scripts/validate_mermaid.py -

It checks the two things you cannot check by eye: state-id consistency (F11) and WCAG AA contrast of every themeVariables foreground/background pair (F13, ≥ 4.5:1 — pure math, no rendering surface needed). Add --cli to also shell out to mmdc for a full parse when the binary is present (fail-open when absent).

Exit 1 means at least one FAIL: fix the diagram and re-validate before emitting. Do not show the user a diagram that failed this check.

Visual layout and communicative quality remain operator-side — hence the Step 5 disclaimer. This step covers only what is mechanically decidable. (Wired 2026-07-26: the validator shipped with a coverage proof but no render skill invoked it.)

Step 5: Append disclaimers

Per engine/render-conventions.md:

• Lossy-on-export (mermaid + ascii + markdown-list): list fields dropped. For ost-render: four_risks, evidence_sources, assumption_tests, prose notes, decision_log_refs.
• Redaction footnote if any anon-labels emitted: list anon-label → registry-entry mapping for audit.
• Carve-out footnote pointers for any literal name whose registry entry has non-empty note:.
• Canonical disclaimer: final block.
• mermaidchart.com handoff for --format mermaid only.

Rules

1. Read-only. Never modify opportunities.yml or any state.
2. If --root-outcome <id> does not exist in opportunities.yml, fail loud; do NOT silently render the full tree.
3. If opportunities.yml is empty, emit a placeholder mindmap with No opportunities yet — run /mycelium:ost-builder + canonical disclaimer. Don't error.
4. Never invent opportunities or solutions not in the canvas.
5. Consent gate is non-skippable. Identifier-bearing fields go through Step 2 regardless of --format. The only override is --no-identifiers=true.

Counter-Argument Check

Before emitting:

1. "Am I rendering opportunities that are archived but still load-bearing in current strategy, or treating archived as deleted?" Default --include-status=all to avoid silent-omission bias.
2. "Did I consult the attribution-registry for EVERY name-shaped identifier in evidence_sources and notes, not just the obvious ones?" The fraud risk is in prose names that look like URLs. Bias toward registry consultation when ambiguous.
3. "Did I confirm cScale* palette is set per engine/render-conventions.md#verified-working-palette?" The mindmap default palette uses rotating bright colors with white text and fails WCAG AA. The verified-working palette is the required default for theme: base.

What this skill does NOT do

• Does NOT build the OST. That's /mycelium:ost-builder.
• Does NOT score solutions. That's /mycelium:ice-score.
• Does NOT cross-cut to cycles or solutions launched. That's the dispatcher's --view traceability (deferred Phase 4a–4d).

Test fixtures (G-V12 / Check 37)

• tests/bash/fixtures/ost-render/empty-canvas.yml → assert placeholder + pointer to /mycelium:ost-builder
• tests/bash/fixtures/ost-render/single-opp-single-sol.yml → assert mindmap with one branch
• tests/bash/fixtures/ost-render/multi-opp-archived-filter.yml → assert --include-status=active excludes archived branches
• tests/bash/fixtures/ost-render/root-outcome-subtree.yml → assert --root-outcome opp-002 renders only that subtree
• tests/bash/fixtures/ost-render/missing-root-outcome.yml → assert fail-loud
• tests/bash/fixtures/ost-render/redaction-public-ok-literal.yml → assert public_ok name renders literally
• tests/bash/fixtures/ost-render/redaction-generic-only-anon.yml → assert generic_only name redacts to anon-label
• tests/bash/fixtures/ost-render/redaction-no-registry-entry-fail-loud.yml → assert fail-loud when name absent from registry
• tests/bash/fixtures/ost-render/all-formats.yml → assert mermaid, ascii, markdown-list, json all emit valid output

Theory citations

• Torres (OST shape, evidence-only)
• Hick's Law (single recommended format default; explicit fail-loud on unsupported)
• WCAG 2.1 AA (mindmap palette per engine/render-conventions.md#wcag-aa-theme-convention)