wiki

> Ported from oh-my-codex `wiki`. OMX runtime conventions (`$macro` invocation, > `omx` CLI, `.omx/` state directory) are replaced with WorkBuddy idioms > (Skill tool, Agent tool, task list, `.workbuddy/memory`). # Wiki Persistent, self-maintained markdown knowledge base for project and session knowledge. There is no `omx` binary; every operation below is implemented with the WorkBuddy file tools (Write, Read, Edit, Grep, Glob, Bash). The wiki lives under `.omw/wiki/` so it is co-located with other WorkBuddy artifacts and easy to gitignore or commit as desired. ## Storage layout - Pages: `.omw/wiki/<slug>.md` - Index: `.omw/wiki/index.md` - Log: `.omw/wiki/log.md` `<slug>` is a kebab-case page id derived from the title (e.g. `auth-architecture`). ## Categories `architecture`, `decision`, `pattern`, `debugging`, `environment`, `session-log`, `reference`, `convention` ## Operations ### Add / Ingest a page 1. Derive the slug from the title. 2. Write `.omw/wiki/<slug>.md` with this shape: ```markdown --- title: <Title> category: <category> tags: [tag1, tag2] created: <YYYY-MM-DD> updated: <YYYY-MM-DD> --- <content in markdown> ``` 3. Append a one-line entry to `.omw/wiki/index.md`: `- [<Title>](<slug>.md) — <category> — tags: tag1, tag2` 4. Append to `.omw/wiki/log.md`: `<YYYY-MM-DD> ADD <slug> (<category>)`. Use the Write tool for new pages. For an existing page, use Edit to update `content` and bump `updated:`. ### Query Keyword + tag search (no vector embeddings — exact substring/tag match only): - By tag/category: `Grep` for `tags:.*<tag>` or `category: <category>` across `.omw/wiki/`. - By keyword: `Grep` for the phrase across `.omw/wiki/*.md` (use `output_mode: files_with_matches`). - Read `index.md` first for a fast directory of all pages. - When multiple candidates match, open the most relevant page with Read and summarize to the user. ### Lint Check the wiki for consistency: - Every page under `.omw/wiki/*.md` (except `index.md`/`log.md`) has valid frontmatter (`title`, `category`, `tags`). - Every `index.md` entry points to a page that exists; every page is listed in `index.md`. - `[[page-name]]` wiki-links resolve to an existing `<slug>.md`. - Report or fix (via Edit) any orphan pages, missing index entries, or broken links. ### Read Open `.omw/wiki/<slug>.md` with Read and present the content. Resolve any `[[page-name]]` cross-references by linking to the corresponding `<slug>.md`. ### Delete 1. Delete `.omw/wiki/<slug>.md` (Bash `rm`). 2. Remove its line from `index.md` (Edit). 3. Append to `log.md`: `<YYYY-MM-DD> DELETE <slug>`. ### Refresh Rebuild `index.md` from the current set of pages (useful after many manual edits): list `.omw/wiki/*.md`, parse each page's frontmatter, and regenerate the index entries. ## Cross-References Use `[[page-name]]` wiki-link syntax inside page content to create cross-references between pages. During lint/read, resolve `[[page-name]]` to the matching `<slug>.md` (match on title or slug). ## Auto-Capture At session end, you may capture discoveries as `session-log-*` pages (category `session-log`) following the Add steps above. Do this proactively when the user learned something durable (debugging fix, environment quirk, decision rationale). No config file is required — just follow the Add workflow. ## Hard Constraints - No vector embeddings — query uses keyword + tag matching only. - Wiki files are repository project knowledge under `.omw/wiki/`; treat them as the canonical knowledge base for the project.