Manage Fux corpus sources with add, remove, and ingest commands.
npx skills add https://github.com/arpitarya/fux --skill fux-sources-arpitarya
Please help me install this Agent Skill. Skill: fux-sources Source: https://github.com/arpitarya/fux/tree/main/.agents/skills/fux-sources Command: npx skills add https://github.com/arpitarya/fux --skill fux-sources-arpitarya
.fux/ correctly, and mistakes like stale exclusions or wrong attributes silently leave documents unindexed. ## Core Features & Use Cases - Corpus Editing: Add or remove directories, files, file-type patterns, and URLs with fux add and fux remove, including archived, keep, ttl, meta, and update attributes. - Re-ingestion Control: Run fux ingest in its correct form — --refetch-all, --no-fetch, --check, --failed — since fux update was removed in 3.0. - Diagnosing Missing Files: Answer "why isn't file X indexed" using fux ingest --list-skipped and the skip-reason table covering type allowlists, ! exclusions, and .fuxignore lines. - Use Case: A teammate asks you to index the docs/runbooks folder and a Confluence page, then refresh stale URLs; this Skill gives the exact commands and warns that a bare fux ingest hits the network. ## Quick Start Ask the assistant to index a folder or URL in the Fux corpus, for example by saying "add docs/runbooks to the fux index and re-ingest".High-intent search queries and answers about installing and using this skill.
Run `fux add docs/runbooks` for a directory or `fux add https://...` for a URL; the entry's scheme decides which list in `.fux/sources/` it joins. Each add ends in an ingest unless you pass `--no-ingest`, and `--dry-run` previews the change without writing.
Use `fux ingest` for every re-ingest, since `fux update` was deleted in 3.0 and now fails as an invalid choice. A bare ingest fetches URLs known to be stale, `--refetch-all` fetches every URL, and `--no-fetch` works fully offline.
Run `fux ingest --list-skipped` to get one `path: reason` per line, covering type allowlist rejections, `!` exclusions, `.fuxignore` hits, and binary or non-UTF8 files. Note that ingest writes skip lines into `.fuxignore`, so after fixing the cause you must delete that stale line and re-ingest.
No, `fux ingest` exits 0 even when fetches fail, keeping the prior record and printing `! <url> — <reason>` on stderr. You must read stderr rather than trusting the exit code, and use `fux remove` to drop a dead page.
Run `fux remove X --dry-run` first: if X has its own line it is deleted, otherwise a `!X` exclusion line is added to `.fux/sources/dirs`. Undoing an exclusion requires deleting both the `!` line and the generated lines in `.fux/.fuxignore`.
api-evangelistManage sharing permissions for tables, workflows, and agents via the Frontline CLI.
Boise-State-DevelopmentSearch the web for current information and cite source URLs.
MercurykzAutomates searching, creating, updating, and commenting on Confluence pages and spaces.
mistakeknotReplace AI agent spinner verbs with Culture ship names from a customizable text file.