SkillsAdd

karlorz/llm-wiki · Archived

wiki-adapter-prd

Map foreign PRD formats (CodeStable, RFCs, structured markdown) into skillwiki raw + typed-knowledge pages.

First seen Jul 27, 2026

Installation

$ npx skills add karlorz/llm-wiki --skill wiki-adapter-prd
View on skills.sh ↗ GitHub repository ↗

Stronger alternatives

This repository is archived — consider an actively maintained alternative.

vault-sync-install Same source
karlorz/llm-wiki

Install vault-sync on the current host. Detects OS, deploys scripts, registers scheduler jobs, …

1 installs
vault-sync-uninstall Same source
karlorz/llm-wiki

Remove vault-sync from current host. Stops scheduler jobs, leaves tombstone markdown, preserves…

1 installs
vault-sync-status Same source
karlorz/llm-wiki

Health snapshot of vault-sync — scheduler health, push/fetch recency, filter integrity, snapsho…

1 installs
vault-fuse-freshness Same source
karlorz/llm-wiki

Audit and refresh rclone FUSE visibility freshness for cross-device wiki sync. Linux-focused; e…

1 installs

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

prisma-driver-adapter-implementation Similar name
prisma/skills

Required reference for Prisma ORM 7 SQL driver adapter work.

258.3K installs
opencli-adapter-author Similar name
jackwener/opencli

Use when writing an OpenCLI adapter for a new site or adding a new command to an existing site.

14.9K installs
mantis-pipeline-adapter Similar name
google/mantis

>- Interactively guides the design and implementation of custom deterministic orchestrator harn…

39 installs
add-adapter Similar name
vercel/chat

Add a vendor-official or community adapter to the Chat SDK catalog and docs — adapters.json reg…

590 installs
create-agent-adapter Similar name
paperclipai/paperclip

Create or modify Paperclip agent adapters across server, UI, and CLI surfaces. Use when adding …

252 installs
npu-adapter-reviewer Similar name
ascend/agent-skills 127 installs

Also in this package

Other skills from karlorz/llm-wiki · top by installs.

npx skills add karlorz/llm-wiki
vault-sync-install 1 vault-sync-uninstall 1 vault-sync-status 1 vault-fuse-freshness 1 wiki-gate-plan-mode 1 vault-presync 1 vault-snapshot 1 wiki-sync 1 wiki-remove 1 wiki-reingest 1

Browse all from karlorz/llm-wiki

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Repository health

Stars 3
Default branch main
Open issues 1
Status Archived

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 4,667 B
  • docs SUMMARY.md 131 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 1 installs

SKILL.md

wiki-adapter-prd

When This Skill Activates

  • User provides a document or URL in a non-skillwiki PRD format and wants it captured in the vault.
  • User mentions CodeStable, RFC, AIDE, or another structured design document format.
  • A foreign spec/plan needs to be normalized into the vault's raw + concept structure.

Output language

Run skillwiki lang at the start. Generate page prose in the resolved language. Frontmatter keys, file names, and structural markers stay English.

Pre-orientation reads

Standard four reads (SCHEMA, index, log, project context if applicable).

Recognized PRD Formats

Format Structural cues
CodeStable REQ-NNN requirement IDs, ## Requirements / ## Architecture headers
RFC ## Motivation / ## Proposal / ## Drawbacks headers
AIDE directives Specific YAML frontmatter keys (aide-*)
Hermes spec N1–N18 normative requirement markers
Generic structured Clear ## section hierarchy with requirements, decisions, or designs

If the format is unrecognized, treat as generic structured markdown and map by section hierarchy.

Mapping Strategy

Raw capture (verbatim)

  • Write the full source document to raw/articles/<slug>.md with RawSourceSchema frontmatter (sha256, sourceurl, ingested, ingestedby: "wiki-ingest").
  • If the source is a URL, run skillwiki fetch-guard <url> first.
  • Run skillwiki hash <raw-file> to compute sha256.

Knowledge extraction

Map source sections to typed-knowledge pages:

Source section Target type Notes
Requirements list concepts/ or entities/ Each major requirement becomes its own page or a section in a compound page
Architecture decisions concepts/ Map to concept pages with tags: [architecture]
Motivation / context entities/ Capture as entity pages describing the system or component
Trade-offs / comparisons comparisons/ Create comparison pages when the source weighs alternatives; include a Decision Closeout block
Action items / next steps Skip Not knowledge — track in project work items instead

Cross-reference handling

  • Requirement IDs (REQ-NNN, N1–N18) → preserve as frontmatter tags or inline references.
  • Internal links within the source → convert to [[wikilinks]] where corresponding pages exist.
  • External URLs → keep as-is in body text.

Steps

  1. Resolve vault and language: skillwiki path and skillwiki lang.
  2. Classify the input format using the structural cues above.
  3. If URL source: run skillwiki fetch-guard <url>, then fetch.
  4. Sensitive content guard. Before writing the raw capture or generated pages, scan the source and generated body for live credentials, access keys, tokens, passwords, cookies, bearer headers, or private keys. Redact generated prose before writing. If the source itself contains a live secret and would need to remain raw, STOP instead of preserving it.
  5. Write raw capture: frontmatter + full body → raw/articles/<slug>.md.
  6. Run skillwiki hash <raw-file>, embed sha256.
  7. Generate typed-knowledge pages following the mapping strategy.

For generated comparison or evaluation pages, end the body with: ```markdown ## Decision Closeout

Disposition: no-op | concept | ADR | work-item | evidence-needed Reason: ... Follow-up: ... `` Use exactly one disposition. Preserve action items as skipped project-management content unless the closeout explicitly says work-item`.

  1. For each page: run skillwiki validate <page>. If any fails, STOP.
  2. Write pages, then update index.md and log.md.

Provenance defaults

  • provenance: research (external PRD sources).
  • sources: ["^[raw/articles/<slug>.md]"] on every generated page.

Stop conditions

  • fetch-guard non-zero.
  • validate non-zero on any page.
  • sha256 already exists for the same source (skip — already ingested).
  • Source or generated content contains unredacted live credentials or other authenticating secrets.

Forbidden

  • Skipping fetch-guard for URL sources.
  • Writing index/log before all pages validate.
  • Modifying existing raw files (N9).
  • Auto-generating pages for action items, timelines, or process steps — those are project management, not knowledge.
  • Writing live credentials, access keys, tokens, passwords, cookies, bearer headers, private keys, or other authenticating secrets to the vault.