wiki_import

wiki_import

<wiki_import> Resource-to-wiki reconciliation editor. Capture the resource the user pointed to as raw source, mine the captured raw for durable knowledge, diff against the wiki, and triage the result before any wiki-page write. Pull a single named external resource into the wiki end to end: persist the raw, classify each candidate by page type and relation to existing content, surface conflicts between the resource and existing wiki pages, and propose reconciliations the user can accept page by page. Defer all wiki structure, discovery, orientation, raw capture, ingest, and lint behavior to the wiki skill. <resolve_first>Resolve $WIKI through the wiki skill's discovery flow before reading the resource. Honor exit-2 ambiguity by presenting candidates and asking the user.</resolve_first> <orient_first>Read SCHEMA.md, index.md, and roughly the last 350 lines of log.md before diffing, so the diff runs against an understood corpus and domain.</orient_first> <confirm_resource>Confirm the resource pointer (URL, file path, paste, PDF, transcript, meeting note, internal note) and its kind with the user before fetching. For a URL to an externally-published article, confirm the target slug under raw/articles/; for a PDF or paper, under raw/papers/; for a meeting note, interview, or spoken-word transcript, under raw/meetings/; for an internal memo, discussion writeup, ad-hoc observation, or internal doc not published externally, under raw/notes/; for a paste, the appropriate raw/ subdirectory by kind. For edge cases (article that embeds a transcript, transcript of a private meeting, paste of unknown provenance, file that fits two buckets equally), consult $WIKI_SKILL/references/raw_taxonomy.md — the canonical reference for bucket meanings and classification heuristics.</confirm_resource> <capture_raw>Route the resource through the wiki skill's Ingest §1 — fetch and convert (URL), extract (PDF), or file (paste, meeting, note) — into raw/<kind>/<slug>.md with the required frontmatter (source_url, ingested, body-only sha256). Write the sha256 with python3 $WIKI_SKILL/scripts/compute_sha256.py raw/<kind>/<slug>.md — never invent the value by hand. On re-ingest of the same URL: rerun the same command, skip if it reports ok, flag drift if it reports update.</capture_raw> <mine_resource>Read the captured raw end to end. Extract durable claims, decisions, definitions, conventions, comparisons, workflows, and named entities. Skip passing mentions, minor details, and material outside the wiki's stated domain in SCHEMA.md.</mine_resource> For each candidate, pick a page type from the wiki skill's enum and tag it NEW, EXTEND, CONFIRM, or CONFLICT against $WIKI. Honor the page-threshold rules from the wiki skill — a passing mention does not earn a page. <surface_contradictions>For every CONFLICT, capture both excerpts verbatim, name the disagreement dimension (factual / definitional / scope / recency / source-quality), and offer two or three concrete reconciliation paths.</surface_contradictions> <propose_then_act>Emit the proposal and wait for user selections before any wiki-page write. The raw capture in step 4 stays — only the entity, concept, comparison, summary, query, and procedure writes wait.</propose_then_act> <defer_writes>Route approved NEW and EXTEND items through the wiki skill's Ingest flow (steps 3–6). Route approved CONFLICT items through its contested-page protocol (contested: true, contradictions: frontmatter, both positions recorded with dates and sources).</defer_writes> <single_resource>Source material for the proposal is the captured raw only. Skip claims the resource did not establish, even when they are true in the broader domain.</single_resource> Confirm the resource pointer and its kind. If the user supplied only "this link" or "that paper", request the URL or path explicitly before continuing. Run discover_wiki.sh and resolve $WIKI. On exit 2, present candidates in walk order and stop until the user picks. Read SCHEMA.md, index.md, and roughly the last 350 lines of log.md. Capture the resource into raw/<kind>/<slug>.md per the wiki skill's Ingest §1 (frontmatter, body-only sha256 via python3 $WIKI_SKILL/scripts/compute_sha256.py, drift check on re-ingest). Read the captured raw end to end. Build a list of durable items, each with title, suggested page type, target slug, and an excerpt anchor into the captured raw. Search $WIKI for each item (index plus recursive grep for 100+-page wikis). Tag NEW, EXTEND, CONFIRM, or CONFLICT. For CONFLICT items, capture both excerpts verbatim. Emit one report under three H2 headings: ## New pages, ## Extensions to existing pages, ## Contradictions to reconcile. Skip CONFIRM entries unless the user asked for a full inventory. On approval, route each item through the wiki skill's matching flow, run python3 scripts/lint.py, and append a single ## [YYYY-MM-DD] import | Source Title — N new, N extended, N contested entry to log.md listing only files actually created or updated. <output_contract> <proposal_format>Three H2 sections. Each entry: title, suggested type, target path, excerpt anchor into the captured raw. Contradictions also carry wiki excerpt, raw excerpt, disagreement dimension, and ≥2 reconciliation options.</proposal_format> <no_wiki_page_writes_before_approval>Skip every entity, concept, comparison, summary, query, and procedure write during the proposal phase. The raw capture is the only write that lands before approval.</no_wiki_page_writes_before_approval> <log_entry>One dated import entry that lists only files actually created or updated (raw plus wiki pages). Skip files inspected, considered, or deliberately left unchanged.</log_entry> Every candidate carries type and target; every CONFLICT carries both excerpts and ≥2 reconciliation options; no wiki page is modified before approval; the log entry matches actual file changes. </output_contract> </wiki_import>