Chorus

 view release on metacpan or  search on metacpan

agent/skills/chorus-feed.md  view on Meta::CPAN

**Extension case:**
- Update `Frame catalog` (new slots)
- Update `Slot dictionary`
- Add rules to `Rule catalog`
- Generate the new YAML files
- Add required helpers to `Helpers.pm`
- Verify that new slots do not conflict with those
  of other agents (Slot dictionary of the index)

**New domain case:**
- Apply Mode A (Phases 1 to 5.5) on the fragment only
- Determine the position of the new agent in the pipeline:
  - Does it read a slot set by an existing agent? → after it
  - Does it set a slot consumed by an existing agent? → before it
- Update `index.org`: insert the new agent at the correct position
- âš  Verify that the insertion does not break the chain of targeting slots

### Phase B4 — Enrichment closing

Update `README.org`:
- Add the row in `Corpus` (number + file + source + date)

agent/skills/chorus-pdf.md  view on Meta::CPAN

TEXT
- Extract all text in reading order (top to bottom, left to right).
- For multi-column layouts: extract column 1 fully, then column 2. Insert a blank line between columns.
- Preserve section numbers, article numbers, and clause references exactly as printed.
- Preserve all footnote markers and footnote text (append footnotes at end of page output).
- Do not summarize, paraphrase, or omit any text.

TABLES
- Reconstruct every table in Markdown format (pipe syntax).
- Preserve all column headers, row labels, units, and footnote references inside the table.
- If a table spans multiple pages: output the fragment visible on this page; prefix it with
  [TABLE CONTINUED — <table title or number>] if this is a continuation.

FIGURES AND DIAGRAMS
- For every figure, diagram, or illustration: output a block of the form:
    [FIGURE <N> — <title or caption>]
    <Structured description of all visual content:>
    - Labeled dimensions, dimensions with units
    - Named components and their spatial relationships
    - Numerical values visible in or next to the figure
    - Arrows, load paths, connection points, hinge symbols, support symbols

agent/skills/chorus-pdf.md  view on Meta::CPAN

**"Figures are described but values seem invented"**
→ LLMs can hallucinate values in dense technical diagrams. Always cross-check critical
  normative values against the original PDF. Mark uncertain values with
  `# TODO: verify against PDF §<N>` in `Helpers.pm`.

**"Auto mode: a text page was sent to vision unnecessarily"**
→ The threshold `len(text.strip()) > 50` in `classify_pages` may be too low for sparse
  pages (cover pages, blank pages, page numbers only). Increase to `> 200` if needed.

**"Chunk boundaries cut through a table"**
→ The fragment is prefixed `[TABLE CONTINUED — ...]`. `chorus-feed` treats both
  fragments as separate text blocks — rarely an issue for KB extraction.

**"Too slow — 54-page PDF with --auto takes a long time"**
→ Only the vision pages hit the API. If 16/54 pages have figures: 4 chunks × ~30s = ~2 min.
  The text pages (pdfminer) complete in seconds. Total: ~2–3 minutes for a 54-page standard.

**"Script exited with code 2 — nohup required"**
→ The layout analysis detected ≥ 16 figures. The script aborted before any API call.
  Copy the `nohup` command printed to stderr and run it in a terminal:
  ```bash
  nohup python3 $SANDBOX/agent/extract-pdf-<slug>.py > $SANDBOX/corpus/<NNN>-<slug>-vision.md.log 2>&1 &

agent/skills/chorus-strengthen.md  view on Meta::CPAN

## Phase 5 — Enrichment roadmap

### 5.1 Group gaps by corrective action

Group all gaps into three buckets:

| Bucket | Content | Action |
|---|---|---|
| **A — Corpus clarification** | Gaps where the normative source is unclear or ambiguous | Need to re-read the original corpus §, possibly extract a correction text |
| **B — Rule adjustment** | Gaps where the corpus reference is known and the fix is a threshold / CONDITION change | Direct YAML edit (does not require `chorus-feed --enrich`) |
| **C — Missing coverage** | Gaps where the KB has no rule at all for this case | New rule needed → `chorus-feed --enrich` with a targeted corpus fragment |

### 5.2 Enrichment corpus recommendation

For bucket **C** (missing coverage), produce a recommendation:

```
Recommended enrichment corpus
──────────────────────────────
For each missing-coverage gap, draft the normative text fragment that covers it.
This text should be passed to:

  chorus-feed <sandbox-name> <corpus-correctif.txt> --enrich

Suggested corpus-correctif.txt content:
─────────────────────────────────────────
[Gap #N — <type>/<slot>]
<Normative extract or paraphrase covering the missing rule.
 Include: element type, triggering condition, threshold value, rejection reason.>