dsh-aris-panel 0.1.0

package/skills/exa-search/SKILL.md

@@ -0,0 +1,205 @@
+---
+name: exa-search
+description: AI-powered web search via Exa with content extraction. Use when user says "exa search", "web search with content", "find similar pages", or needs broad web results beyond academic databases (arXiv, Semantic Scholar).
+argument-hint: "[search-query-or-url]"
+allowed-tools: Bash(*), Read, Write
+---
+
+# Exa AI-Powered Web Search
+
+Search query: $ARGUMENTS
+
+## Role & Positioning
+
+Exa is the **broad web search** source with built-in content extraction:
+
+| Skill | Best for |
+|------|----------|
+| `/arxiv` | Direct preprint search and PDF download |
+| `/semantic-scholar` | Published venue papers (IEEE, ACM, Springer), citation counts |
+| `/deepxiv` | Layered reading: search, brief, section map, section reads |
+| `/exa-search` | Broad web search: blogs, docs, news, companies, research papers — with content extraction |
+
+Use Exa when you need results beyond academic databases, or when you want content (highlights, full text, summaries) extracted alongside search results.
+
+## Constants
+
+- **EXA_FETCHER** — canonical name `exa_search.py`, resolved per
+  [`shared-references/integration-contract.md`](../shared-references/integration-contract.md) §2
+  (Policy D1 — standalone `/exa-search` has no documented fallback,
+  so unresolved helper terminates with an explicit error).
+- **MAX_RESULTS = 10** — Default number of results to return.
+
+> Overrides (append to arguments):
+> - `/exa-search "RAG pipelines" — max: 5` — top 5 results
+> - `/exa-search "diffusion models" — category: research paper` — research papers only
+> - `/exa-search "startup funding" — category: news, start date: 2025-01-01` — recent news
+> - `/exa-search "transformer" — content: text, max chars: 8000` — full text mode
+> - `/exa-search "transformer" — content: summary` — LLM-generated summaries
+> - `/exa-search "transformer" — domains: arxiv.org,huggingface.co` — domain filter
+> - `/exa-search "https://arxiv.org/abs/2301.07041" — similar` — find similar pages
+
+## Setup
+
+Exa requires the `exa-py` SDK and an API key:
+
+```bash
+pip install exa-py
+```
+
+Set your API key:
+```bash
+export EXA_API_KEY=your-key-here
+```
+
+Get a key from [exa.ai](https://exa.ai).
+
+## Workflow
+
+### Step 1: Parse Arguments
+
+Parse `$ARGUMENTS` for:
+- **query**: The search query (required) or a URL (for `find-similar` mode)
+- **similar**: If present, use `find-similar` mode instead of search
+- **max**: Override MAX_RESULTS
+- **category**: `research paper`, `news`, `company`, `personal site`, `financial report`, `people`
+- **content**: `highlights` (default), `text`, `summary`, `none`
+- **max chars**: Max characters for content extraction
+- **type**: Search type — `auto` (default), `neural`, `fast`, `instant`
+- **domains**: Comma-separated include domains
+- **exclude domains**: Comma-separated exclude domains
+- **include text**: Phrase that must appear in results
+- **exclude text**: Phrase to exclude from results
+- **start date**: ISO 8601 date — only results after this
+- **end date**: ISO 8601 date — only results before this
+- **location**: Two-letter ISO country code
+
+### Step 2: Locate Script
+
+Resolve `$EXA_FETCHER` via the canonical strict-safe chain (see
+[`shared-references/integration-contract.md`](../shared-references/integration-contract.md) §2).
+Policy D1 cascade: there is no native inline fallback for Exa
+(retrieval requires the `exa-py` SDK + API key, which lives in the
+fetcher), so unresolved helper means the SKILL cannot produce its
+primary output — fail with explicit remediation.
+
+```bash
+cd "$(git rev-parse --show-toplevel 2>/dev/null || pwd)" || exit 1
+if [ -z "${ARIS_REPO:-}" ] && [ -f .aris/installed-skills.txt ]; then
+    ARIS_REPO=$(awk -F'\t' '$1=="repo_root"{print $2; exit}' .aris/installed-skills.txt 2>/dev/null) || true
+fi
+if [ -z "${ARIS_REPO:-}" ] && [ -f "$HOME/.aris/repo" ]; then
+    ARIS_REPO=$(cat "$HOME/.aris/repo" 2>/dev/null) || true
+fi
+EXA_FETCHER=".aris/tools/exa_search.py"
+[ -f "$EXA_FETCHER" ] || EXA_FETCHER="tools/exa_search.py"
+[ -f "$EXA_FETCHER" ] || { [ -n "${ARIS_REPO:-}" ] && EXA_FETCHER="$ARIS_REPO/tools/exa_search.py"; }
+[ -f "$EXA_FETCHER" ] || {
+  echo "ERROR: exa_search.py not resolved at .aris/tools/, tools/, \$ARIS_REPO/tools/, or via ~/.aris/repo." >&2
+  echo "       Fix: rerun bash tools/install_aris.sh or smart_update.sh (refreshes ~/.aris/repo), export ARIS_REPO, or copy the helper to tools/." >&2
+  echo "       Also ensure 'exa-py' is installed: pip install exa-py" >&2
+  exit 1
+}
+```
+
+### Step 3: Execute Search
+
+**Standard search:**
+```bash
+python3 "$EXA_FETCHER" search "QUERY" --max 10 --content highlights
+```
+
+**With filters:**
+```bash
+python3 "$EXA_FETCHER" search "QUERY" --max 10 \
+  --category "research paper" \
+  --start-date 2025-01-01 \
+  --content text --max-chars 8000
+```
+
+**Find similar pages:**
+```bash
+python3 "$EXA_FETCHER" find-similar "URL" --max 5 --content highlights
+```
+
+**Get content for known URLs:**
+```bash
+python3 "$EXA_FETCHER" get-contents "URL1" "URL2" --content text
+```
+
+### Step 4: Present Results
+
+Format results as a structured table:
+
+```
+| # | Title | Authors | Venue/Publisher | URL | Date | Key Content |
+|---|-------|---------|-----------------|-----|------|-------------|
+```
+
+For each result:
+- Show title and URL
+- Show published date if available
+- Show highlights, text excerpt, or summary depending on content mode
+- Flag particularly relevant results
+- **For `category: "research paper"` hits only** — also record authors
+  (from Exa's `author`/`authors` fields, or fallback: parse from the
+  result snippet) and venue/publisher (from `publisher`, `source`, or
+  the domain hosting the paper). These are needed by Step 6's wiki
+  hook; if either is unavailable for a given hit, skip wiki ingest
+  for that one hit and log a note.
+
+### Step 5: Offer Follow-up
+
+After presenting results, suggest:
+- **Deepen**: "I can fetch full text for any of these results"
+- **Find similar**: "I can find pages similar to any result"
+- **Narrow**: "I can re-search with domain/date/text filters"
+
+### Step 6: Update Research Wiki (if active, research-paper results only)
+
+**Required when `research-wiki/` exists AND the search returned
+results of `category: "research paper"`**; skip silently otherwise.
+General web results (blog posts, docs, news) are **not** ingested —
+the wiki is for papers only.
+
+When the predicates hold, resolve `$WIKI_SCRIPT` per the canonical
+chain at
+[`shared-references/wiki-helper-resolution.md`](../shared-references/wiki-helper-resolution.md)
+(Variant B — warn-and-skip). For each research paper hit, try to
+recover an arXiv ID from the URL (`arxiv.org/abs/<id>`); if present,
+use `--arxiv-id`. Otherwise fall back to manual metadata:
+
+```bash
+if [ -d research-wiki/ ] and query category was "research paper":
+    cd "$(git rev-parse --show-toplevel 2>/dev/null || pwd)" || exit 1
+    ARIS_REPO="${ARIS_REPO:-$(awk -F'\t' '$1=="repo_root"{print $2; exit}' .aris/installed-skills.txt 2>/dev/null)}"
+    if [ -z "${ARIS_REPO:-}" ] && [ -f "$HOME/.aris/repo" ]; then
+      ARIS_REPO=$(cat "$HOME/.aris/repo" 2>/dev/null) || true
+    fi
+    WIKI_SCRIPT=".aris/tools/research_wiki.py"
+    [ -f "$WIKI_SCRIPT" ] || WIKI_SCRIPT="tools/research_wiki.py"
+    [ -f "$WIKI_SCRIPT" ] || { [ -n "${ARIS_REPO:-}" ] && WIKI_SCRIPT="$ARIS_REPO/tools/research_wiki.py"; }
+    [ -f "$WIKI_SCRIPT" ] || {
+      echo "WARN: research_wiki.py not found; exa-search results delivered, wiki ingest skipped. Fix: bash tools/install_aris.sh or smart_update.sh (refreshes ~/.aris/repo), export ARIS_REPO, or cp <ARIS-repo>/tools/research_wiki.py tools/." >&2
+      WIKI_SCRIPT=""
+    }
+    [ -n "$WIKI_SCRIPT" ] && for each research-paper hit in results:
+        if URL matches arxiv.org/abs/<id>:
+            python3 "$WIKI_SCRIPT" ingest_paper research-wiki/ \
+                --arxiv-id "<id>"
+        else:
+            python3 "$WIKI_SCRIPT" ingest_paper research-wiki/ \
+                --title "<title>" --authors "<authors joined by , >" \
+                --year <year> --venue "<venue or publisher>"
+```
+
+The helper handles slug / dedup / page / index / log — **do not
+handwrite `papers/<slug>.md`**. See
+[`shared-references/integration-contract.md`](../shared-references/integration-contract.md).
+
+## Key Rules
+- Always check that `EXA_API_KEY` is set before searching
+- Default to `highlights` content mode for a good balance of speed and context
+- Use `category: "research paper"` when the user is clearly looking for academic content
+- Use `text` content mode when the user needs full page content
+- Combine with `/arxiv` or `/semantic-scholar` for comprehensive literature coverage

package/skills/experiment-audit/SKILL.md

@@ -0,0 +1,311 @@
+---
+name: experiment-audit
+description: "Audit experiment integrity before claiming results. Uses cross-model review (external reviewer backend) to check for fake ground truth, score normalization fraud, phantom results, and insufficient scope. Use when user says \"审计实验\", \"check experiment integrity\", \"audit results\", \"实验诚实度\", or after experiments complete before writing claims."
+argument-hint: "[experiment-dir-or-results-path]"
+allowed-tools: Bash(*), Read, Write, Edit, Grep, Glob, mcp__codex__codex, mcp__codex__codex-reply, mcp__manual_review__review, mcp__manual_review__review_reply
+---
+
+# Experiment Audit: Cross-Model Integrity Verification
+
+> 🔒 **Do not wrap this skill in `/loop`, `/schedule`, or `CronCreate`.** It is
+> verdict-bearing — it judges experiment integrity. Re-running that verdict on a
+> timer adds no new signal, and a loop that accepts its own output to decide
+> when to stop crosses into self-acquittal (`acceptance-gate.md`). Schedule the
+> *external wait that precedes it* — experiments done → then audit **once**. See
+> [`shared-references/external-cadence.md`](../shared-references/external-cadence.md).
+
+Audit experiment integrity for: **$ARGUMENTS**
+
+## Why This Exists
+
+LLM agents can produce fraudulent experimental results through:
+1. **Fake ground truth** — creating synthetic "reference" from model outputs, then reporting high agreement as performance
+2. **Score normalization** — dividing metrics by the model's own max to get 0.99+
+3. **Phantom results** — claiming numbers from files that don't exist or functions never called
+4. **Insufficient scope** — reporting 2-scene pilots as "comprehensive evaluation"
+
+These are NOT intentional deception — they are failure modes of optimizing agents that lack integrity constraints. This skill adds that constraint.
+
+## Core Principle
+
+**The executor collects file paths. The external reviewer backend reads code and judges integrity. The executor does NOT participate in integrity judgment.**
+
+This follows `shared-references/reviewer-independence.md` and `shared-references/experiment-integrity.md`.
+
+## Constants
+
+- **REVIEWER_BACKEND = `codex`** — Default: Codex MCP (ultra). Override with `— reviewer: oracle-pro` for Oracle MCP, or `— reviewer: manual` for Manual Review MCP. If manual-review MCP is unavailable, stop and print the install command; do not fall back to Codex. See `shared-references/reviewer-routing.md`.
+
+## Reviewer Calling Convention
+
+When calling the reviewer, branch on REVIEWER_BACKEND:
+
+**If REVIEWER_BACKEND = `codex`:**
+  Use `mcp__codex__codex` for new review threads.
+  Use `mcp__codex__codex-reply` for follow-up rounds (reuse threadId).
+
+**If REVIEWER_BACKEND = `manual`:**
+  Use `mcp__manual_review__review` for new review threads with:
+    prompt: [exact same prompt that would go to Codex]
+    config: {"model_reasoning_effort": "xhigh", "executor_model": "<actual executor model>", "require_reviewer_model": true}
+  Save the returned `threadId`.
+  Use `mcp__manual_review__review_reply` for follow-up rounds with:
+    threadId: [saved manual-review threadId]
+    prompt: [follow-up prompt]
+    config: {"model_reasoning_effort": "xhigh", "executor_model": "<actual executor model>", "require_reviewer_model": true}
+
+Prompt fidelity: the manual prompt must be exactly the same text that Codex would receive.
+Review tracing applies equally to both backends.
+
+## Workflow
+
+### Step 1: Collect Artifacts (Executor — Claude)
+
+Locate and list these files WITHOUT reading or summarizing their content:
+
+```
+Scan project directory for:
+1. Evaluation scripts:    *eval*.py, *metric*.py, *test*.py, *benchmark*.py
+2. Result files:          *.json, *.csv in results/, outputs/, logs/
+3. Ground truth paths:    look in eval scripts for data loading (dataset paths, GT references)
+4. Experiment tracker:    EXPERIMENT_TRACKER.md, EXPERIMENT_LOG.md
+5. Paper claims:          NARRATIVE_REPORT.md, paper/sections/*.tex, PAPER_PLAN.md
+6. Config files:          *.yaml, *.toml, *.json configs with metric definitions
+```
+  A verdict-bearing manual response MUST begin with
+  `Reviewer-Model: <exact-model-id>` — pass the model THIS session is actually
+  running as in `executor_model`. Missing, unknown, or same-family identity
+  cannot acquit; emit `REVIEW_UNAVAILABLE` rather than guessing. If the executor
+  model cannot be named, manual review's cross-family claim is unprovable — say
+  so in the report instead of asserting it.
+
+
+**DO NOT summarize, interpret, or explain any file content.** Only collect paths.
+
+### Step 2: Send to Reviewer
+
+Based on the selected reviewer backend (see Reviewer Calling Convention), pass ONLY file paths and the audit checklist to the reviewer. The reviewer reads everything directly.
+
+For `codex`, call `mcp__codex__codex` with:
+- `model: gpt-5.6-sol`
+- `config: {"model_reasoning_effort": "ultra"}`
+- `sandbox: read-only`
+- `cwd: [project directory]`
+- `prompt: [the exact full prompt below]`
+
+For `manual`, call `mcp__manual_review__review` with:
+- `config: {"model_reasoning_effort": "xhigh", "executor_model": "<actual executor model>", "require_reviewer_model": true}`
+- `prompt: [the exact full prompt below]`
+
+Manual review cannot use Codex-only `model`, `sandbox`, or `cwd`; include the same file paths in the prompt so the user can inspect them.
+
+Use this exact prompt for both backends:
+
+```
+You are an experiment integrity auditor. Start from the assumption that the
+    evaluation is compromised somewhere — your job is to find where. Be
+    adversarial. Trust nothing the author tells you — verify everything
+    yourself. Read ALL files listed below and check for the following fraud
+    patterns.
+
+    Files to read:
+    - Evaluation scripts: [list paths]
+    - Result files: [list paths]
+    - Experiment tracker: [list paths]
+    - Paper claims: [list paths]
+    - Config files: [list paths]
+
+    ## Audit Checklist
+
+    ### A. Ground Truth Provenance
+    For each evaluation script:
+    1. Where does "ground truth" / "reference" / "target" come from?
+    2. Is it loaded from the DATASET, or generated/derived from MODEL OUTPUTS?
+    3. If derived: is it explicitly labeled as proxy evaluation?
+    4. Are official eval scripts used when available for this benchmark?
+    FAIL if: GT is derived from model outputs without explicit proxy labeling.
+
+    ### B. Score Normalization
+    For each metric computation:
+    1. Is any metric divided by max/min/mean of the model's OWN output?
+    2. Are raw scores reported alongside any normalized scores?
+    3. Are any scores suspiciously close to 1.0 or 100%?
+    FAIL if: Normalization denominator comes from prediction statistics.
+
+    ### C. Result File Existence
+    For each claim in the paper/narrative:
+    1. Does the referenced result file actually exist?
+    2. Does the claimed metric key exist in that file?
+    3. Does the claimed NUMBER match what's in the file?
+    4. Is the experiment tracker status DONE (not TODO/IN_PROGRESS)?
+    FAIL if: Claimed results reference nonexistent files or mismatched numbers.
+
+    ### D. Dead Code Detection
+    For each metric function defined in eval scripts:
+    1. Is it actually CALLED in any evaluation pipeline?
+    2. Does its output appear in any result file?
+    WARN if: Metric functions exist but are never called.
+
+    ### E. Scope Assessment
+    1. How many scenes/datasets/configurations were actually tested?
+    2. How many seeds/runs per configuration?
+    3. Does the paper use words like "comprehensive", "extensive", "robust"?
+    4. Is the actual scope sufficient for those claims?
+    WARN if: Scope language exceeds actual evidence.
+
+    ### F. Evaluation Type Classification
+    Classify each evaluation as:
+    - real_gt: uses dataset-provided ground truth
+    - synthetic_proxy: uses model-generated reference
+    - self_supervised_proxy: no GT by design
+    - simulation_only: simulated environment
+    - human_eval: human judges
+
+    ## Output Format
+
+    For each check (A-F), report:
+    - Status: PASS | WARN | FAIL
+    - Evidence: exact file:line references
+    - Details: what specifically was found
+
+    Overall verdict: PASS | WARN | FAIL
+    
+    Be thorough. Read every eval script line by line.
+```
+
+### Step 3: Parse and Write Report (Executor — Claude)
+
+Parse the reviewer's response and write `EXPERIMENT_AUDIT.md`:
+
+```markdown
+# Experiment Audit Report
+
+**Date**: [today]
+**Auditor**: External reviewer backend, ultra reasoning (cross-model, read-only)
+**Project**: [project name]
+
+## Overall Verdict: [PASS | WARN | FAIL]
+
+## Integrity Status: [pass | warn | fail]
+
+## Checks
+
+### A. Ground Truth Provenance: [PASS|WARN|FAIL]
+[details + file:line evidence]
+
+### B. Score Normalization: [PASS|WARN|FAIL]
+[details]
+
+### C. Result File Existence: [PASS|WARN|FAIL]
+[details]
+
+### D. Dead Code Detection: [PASS|WARN|FAIL]
+[details]
+
+### E. Scope Assessment: [PASS|WARN|FAIL]
+[details]
+
+### F. Evaluation Type: [real_gt | synthetic_proxy | ...]
+[classification + evidence]
+
+## Action Items
+- [specific fixes if WARN or FAIL]
+
+## Claim Impact
+- Claim 1: [supported | needs qualifier | unsupported]
+- Claim 2: ...
+```
+
+Also write `EXPERIMENT_AUDIT.json` for machine consumption:
+
+```json
+{
+  "date": "2026-04-10",
+  "auditor": "external-reviewer-ultra",
+  "overall_verdict": "warn",
+  "integrity_status": "warn",
+  "checks": {
+    "gt_provenance": {"status": "pass", "details": "..."},
+    "score_normalization": {"status": "warn", "details": "..."},
+    "result_existence": {"status": "pass", "details": "..."},
+    "dead_code": {"status": "pass", "details": "..."},
+    "scope": {"status": "warn", "details": "..."},
+    "eval_type": "real_gt"
+  },
+  "claims": [
+    {"id": "C1", "impact": "supported"},
+    {"id": "C2", "impact": "needs_qualifier"}
+  ]
+}
+```
+
+### Step 4: Print Summary
+
+```
+🔬 Experiment Audit Complete
+
+  GT Provenance:      ✅ PASS — real dataset GT used
+  Score Normalization: ⚠️ WARN — boundary metric uses self-reference
+  Result Existence:    ✅ PASS — all files exist, numbers match
+  Dead Code:           ✅ PASS — all metric functions called
+  Scope:               ⚠️ WARN — 2 scenes, paper says "comprehensive"
+
+  Overall: ⚠️ WARN
+  
+  See EXPERIMENT_AUDIT.md for details.
+```
+
+## Integration with Other Skills
+
+### Automatic in /research-pipeline (advisory, never blocks)
+
+When integrated into the pipeline, this skill runs automatically after `/experiment-bridge` and before `/auto-review-loop`:
+
+```
+/experiment-bridge → results ready
+    ↓
+/experiment-audit (automatic, advisory)
+    ├── PASS  → continue normally
+    ├── WARN  → print ⚠️ warning, continue, tag claims as [INTEGRITY: WARN]
+    └── FAIL  → print 🔴 alert, continue, tag claims as [INTEGRITY CONCERN]
+    ↓
+/auto-review-loop → proceeds with integrity tags visible to reviewer
+```
+
+**Never blocks the pipeline.** Even on FAIL, the pipeline continues — but claims carry visible integrity tags.
+
+### Read by /result-to-claim (if exists)
+
+```
+if EXPERIMENT_AUDIT.json exists:
+    read integrity_status
+    attach to verdict: {claim_supported: "yes", integrity_status: "warn"}
+    if integrity_status == "fail":
+        downgrade verdict display: "yes [INTEGRITY CONCERN]"
+else:
+    verdict as normal, integrity_status = "unavailable"
+    mark as "provisional — no integrity audit"
+```
+
+### Read by /paper-write (if exists)
+
+```
+if EXPERIMENT_AUDIT.json exists AND integrity_status == "fail":
+    add footnote to affected claims: "Note: integrity audit flagged concerns with this evaluation"
+```
+
+## Key Rules
+
+- **Reviewer independence**: executor collects paths, reviewer judges. Period.
+- **Never block**: warn loudly, never halt the pipeline.
+- **File-as-switch**: no EXPERIMENT_AUDIT.md = skill was never run = zero impact on existing behavior.
+- **Cross-model**: the reviewer MUST be a different model family from the executor.
+- **Honest about limits**: the audit catches common patterns, not all possible fraud. It is a safety net, not a guarantee.
+
+## Acknowledgements
+
+Motivated by community-reported integrity issues (#57, #131) where executor agents created fake ground truth and self-normalized scores.
+
+## Review Tracing
+
+After each reviewer call (`mcp__codex__codex`, `mcp__codex__codex-reply`, `mcp__manual_review__review`, or `mcp__manual_review__review_reply`), save the trace following `shared-references/review-tracing.md` (Policy C — forensic; never silently skip). Use `save_trace.sh` (resolved per the chain in `shared-references/integration-contract.md` §2) or write files directly to `.aris/traces/<skill>/<date>_run<NN>/`. Respect the `--- trace:` parameter (default: `full`).