Research Wiki: Persistent Research Knowledge Base

Subcommand: $ARGUMENTS

Overview

The research wiki is a persistent, per-project knowledge base that accumulates structured knowledge across the entire Prism research lifecycle. Unlike one-off literature surveys that are used and forgotten, the wiki compounds — every paper read, idea tested, experiment run, and review received makes the wiki smarter.

Inspired by Karpathy's LLM Wiki pattern: compile knowledge once, keep it current, don't re-derive on every query.

Core Concepts

Four Entity Types

Typed Relationships (graph/edges.jsonl)

Edges are stored in graph/edges.jsonl only. The ## Connections section on each page is auto-generated from the graph — never hand-edit it.

Wiki Directory Structure

research-wiki/
  index.md               # categorical index (auto-generated)
  log.md                 # append-only timeline
  gap_map.md             # field gaps with stable IDs (G1, G2, ...)
  query_pack.md          # compressed summary for /research-refine (auto-generated, max 8000 chars)
  papers/
    <slug>.md            # one page per paper
  ideas/
    <idea_id>.md         # one page per idea
  experiments/
    <exp_id>.md          # one page per experiment
  claims/
    <claim_id>.md        # one page per testable claim
  graph/
    edges.jsonl          # materialized current relationship graph

Subcommands

/research-wiki init

Initialize the wiki for the current project:

1. Create research-wiki/ directory structure
2. Create empty index.md, log.md, gap_map.md
3. Create empty graph/edges.jsonl
4. Log: "Wiki initialized"

/research-wiki ingest "<paper title>" — arxiv: <id>

Add a paper to the wiki. This subcommand is thin wrapping around the canonical helper python3 tools/research_wiki.py ingest_paper …, which is the single implementation of paper ingest in Prism (per shared-references/integration-contract.md — one helper, no copies). The helper does all of:

1. Fetch metadata — queries the arXiv Atom API when --arxiv-id is given
2. Generate slug — <first_author_last_name><year>_<keyword>
3. Check dedup — skip an existing page unless --update-on-exist
4. Create page — papers/<slug>.md with the schema below
5. Rebuild index.md and query_pack.md
6. Append log.md

Edge extraction (step 5/8 in the old manual flow) is not in ingest_paper; do it as a follow-up with add_edge per relationship identified:

# arXiv-known paper
python3 tools/research_wiki.py ingest_paper research-wiki/ \
    --arxiv-id 2501.12345 --thesis "One-line claim from abstract."

# Venue paper with no arXiv mirror
python3 tools/research_wiki.py ingest_paper research-wiki/ \
    --title "Attention Is All You Need" \
    --authors "Ashish Vaswani, Noam Shazeer, …" --year 2017 --venue "NeurIPS"

# Manual edge after ingest
python3 tools/research_wiki.py add_edge research-wiki/ \
    --from "paper:vaswani2017_attention_all_you" \
    --to "paper:chen2025_factorized_gap" \
    --type "extends" --evidence "Section 3.2: adapts the encoder block …"

Other skills (/research-lit, /semantic-scholar, /exa-search) call the same helper directly in their own last step — they don't re-route through /research-wiki ingest as a subcommand, so they don't need an LLM roundtrip.

Local PDF ingest (born-digital vs. hard PDFs)

For a paper you have as a local PDF (no arXiv/venue metadata to fetch), the default path is to read the PDF with the agent's own multimodal Read tool and supply --title/--authors/--year/--venue/--thesis to ingest_paper. This needs no dependencies and works for born-digital text.

For the hard cases — scanned court judgments, multi-column legal/academic layouts, table-dense pages, formula-heavy text — agent-native reading can mis-order columns or drop tables. Prism ships an optional bridge to MinerU that converts such PDFs to clean Markdown first:

# Probe once; exit 3 means MinerU is not installed -> just read the PDF natively.
python3 tools/pdf_extract.py --check

# Convert a hard PDF to Markdown, then read that .md instead of the raw PDF.
python3 tools/pdf_extract.py "path/to/judgment.pdf" --print > /tmp/judgment.md

MinerU is not a Prism dependency. The script falls back cleanly (exit code 3) when it is absent, so the zero-dependency core is untouched. Install only if you need it: pip install mineru. Use it selectively for scanned / multi-column / table- or formula-heavy PDFs — for ordinary born-digital papers, native reading is faster and good enough.

/research-wiki sync — arxiv-ids <id1>,<id2>,...

Batch backfill: ingest one or more arXiv IDs that were read earlier without being ingested (e.g., because research-wiki/ was set up after the reading happened, or a hook didn't fire).

# Explicit list
python3 tools/research_wiki.py sync research-wiki/ \
    --arxiv-ids 2310.06770,1706.03762

# From a file (one id per line, # comments ok)
python3 tools/research_wiki.py sync research-wiki/ --from-file ids.txt

Dedup is handled per-id; already-ingested papers are skipped silently. This is the recommended manual repair step (see integration contract §5 Backfill). sync does not scan session traces — callers declare the ids explicitly.

Paper page schema (exactly what ingest_paper emits — do not handwrite alternative fields; lint will flag drift):

---
type: paper
node_id: paper:<slug>
title: "<full title>"
authors: ["First A. Author", "Second B. Author"]
year: 2025
venue: "arXiv"
external_ids:
  arxiv: "2501.12345"
  doi: null
  s2: null
tags: ["tag1", "tag2"]
added: 2026-04-07T10:12:00Z
---

# <full title>

## One-line thesis

[Single sentence capturing the paper's core contribution]

## Problem / Gap

## Method

## Key Results

## Assumptions

## Limitations / Failure Modes

## Reusable Ingredients

[Techniques, datasets, or insights that could be repurposed]

## Open Questions

## Claims

[Reference claim pages: claim:C1, claim:C2, etc.]

## Connections

[AUTO-GENERATED from graph/edges.jsonl — do not edit manually]

## Relevance to This Project

[Why this paper matters for our specific research direction]

Additionally, when the paper was ingested via --arxiv-id and the arXiv API returned an abstract, the helper appends an ## Abstract (original) section after Relevance to This Project containing the raw abstract text as a blockquote. Manual ingests (no --arxiv-id) do not include this section.

/research-wiki query "<topic>"

Generate query_pack.md — a compressed, context-window-friendly summary:

Fixed budget (max 8000 chars / ~2000 tokens):

Pruning priority (when over budget): low-ranked papers > cluster detail > chain detail. Never prune failed ideas or top gaps first.

Key rule: Read from short fields only (frontmatter, one-line thesis, gap summary, failure note). Do not summarize full page bodies every time.

/research-wiki update <node_id> — <field>: <value>

Update a specific entity:

/research-wiki update paper:chen2025 — relevance: core
/research-wiki update idea:001 — outcome: negative
/research-wiki update claim:C1 — status: invalidated

After any update: rebuild query_pack.md, update log.md.

/research-wiki lint

Health check the wiki:

1. Orphan pages — entities with zero edges
2. Stale claims — claims with status: reported older than 14 days
3. Contradictions — claims with both supports and invalidates edges
4. Missing connections — papers sharing 2+ tags but no explicit relationship
5. Dead ideas — stage: proposed ideas that were never tested
6. Sparse pages — pages with 3+ empty sections

Output a LINT_REPORT.md with suggested fixes.

/research-wiki stats

Quick overview:

📚 Research Wiki Stats
Papers: 28 (12 core, 10 related, 6 peripheral)
Ideas: 7 (2 active, 3 failed, 1 partial, 1 succeeded)
Experiments: 12
Claims: 15 (5 supported, 3 invalidated, 7 reported)
Edges: 64
Gaps: 8 (3 unresolved)
Last updated: 2026-04-07T10:12:00Z

Integration with Existing Workflows

All paper-reading skills follow the same integration contract (see shared-references/integration-contract.md):

• single predicate — [ -d research-wiki/ ]
• single canonical helper — python3 tools/research_wiki.py ingest_paper …
• concrete artifact — papers/<slug>.md + log.md entry
• backfill — sync --arxiv-ids …
• diagnostic — tools/verify_wiki_coverage.sh

Hook 1: After /research-lit finds papers

# At end of research-lit, after synthesis:
if research-wiki/ exists:
    for paper in top_relevant_papers (limit 8-12):
        python3 tools/research_wiki.py ingest_paper research-wiki/ \
            --arxiv-id <id> [--thesis "..."] [--tags "..."]
        for each explicit relation to existing wiki paper:
            python3 tools/research_wiki.py add_edge research-wiki/ \
                --from "paper:<slug>" --to "<target>" \
                --type <extends|contradicts|addresses_gap|...> \
                --evidence "..."
    log "research-lit ingested N papers"

Each paper-reading skill ships its own Step "Update Research Wiki (if active)" that calls the same helper once per paper it touched. The business logic is not duplicated — only the loop over that skill's specific result set differs.

Hook 2: /research-refine reads AND writes wiki

Before research-question formulation:

if research-wiki/query_pack.md exists (and < 7 days old):
    prepend query_pack to landscape context
    treat failed questions/theses as banlist
    treat top gaps as search seeds
    still run fresh literature search for foundational + recent works

After formulation (THIS IS CRITICAL — without it, ideas/ stays empty):

for question in all_candidate_questions (kept + rejected):
    /research-wiki upsert_idea(question)
    for source_id in question.based_on:
        add_edge(question.node_id, source_id, "inspired_by")
    for gap_id in question.target_gaps:
        add_edge(question.node_id, gap_id, "addresses_gap")
rebuild query_pack
log "research-refine wrote N candidate questions to wiki"

Hook 3: After /argument-stress-test verdict

# Create experiment page
exp_id = upsert_experiment(experiment_data)

# Update each claim's status
for claim_id in resolved_claims:
    if verdict == "yes":
        set_claim_status(claim_id, "supported")
        add_edge(exp_id, claim_id, "supports")
    elif verdict == "partial":
        set_claim_status(claim_id, "partial")
        add_edge(exp_id, claim_id, "supports")  # partial
    else:
        set_claim_status(claim_id, "invalidated")
        add_edge(exp_id, claim_id, "invalidates")

# Update idea outcome
update_idea(active_idea_id, outcome=verdict)

# If failed, record WHY for future ideation
if verdict in ("no", "partial"):
    update_idea failure_notes with specific metrics and reasons

rebuild query_pack
log "argument-stress-test: exp_id updated, verdict=..."

Re-ideation Trigger

After significant wiki updates, suggest re-running /research-refine:

• ≥5 new sources ingested since last formulation
• ≥3 new failed/partial questions since last formulation
• New contradiction discovered in the graph
• New gap identified that no existing idea addresses

The system suggests but does not auto-trigger. User decides.

Key Rules

• One source of truth for relationships: graph/edges.jsonl. Page Connections sections are auto-generated views.
• Canonical node IDs everywhere: paper:<slug>, idea:<id>, exp:<id>, claim:<id>, gap:<id>. Never use raw titles or inconsistent shorthands.
• Failed ideas are the most valuable memory. Never prune them from query_pack.
• query_pack.md is hard-budgeted at 8000 chars. Deterministic generation, not open-ended summarization.
• Append to log.md for every mutation. The log is the audit trail.
• Reviewer independence applies. When the wiki is read by cross-model review skills, pass file paths only — do not summarize wiki content for the reviewer.

Acknowledgements

Inspired by Karpathy's LLM Wiki — "compile knowledge once, keep it current, don't re-derive on every query."