extraction-form

Name: extraction-form
Author: WILLOSCAR

489stars39forks

How to install extraction-form?

willoscar/research-units-pipeline-skills/extraction-form

$npx -y skills add WILLOSCAR/research-units-pipeline-skills --skill extraction-form

Installs into the current project.

Summary

Use when evidence-review has screened includes and needs a schema-aligned extraction table. Trigger: extraction form, extraction table, data extraction, 信息提取, 提取表. Use when: evidence-review 在 screening 后进入 extraction（C4），需要把纳入论文按字段落到 CSV 以支持后续 synthesis。

Files · 1

View on GitHub

SKILL.md

1	# Extraction Form
2
3	Transforms screened include rows plus protocol schema into the analysis table used by `evidence-review`.
4
5	## Inputs
6
7	Required:
8	- `papers/screening_log.csv`
9	- `output/PROTOCOL.md`
10
11	Optional:
12	- `papers/paper_notes.jsonl`
13
14	## Output
15
16	- `papers/extraction_table.csv`
17
18	## Contract
19
20	The table must:
21	- contain one row per included paper
22	- preserve provenance columns (`paper_id`, `title`, `year`, `url`)
23	- include protocol-defined extraction fields
24	- keep narrative residue in `notes`, not in schema columns
25
26	## Script boundary
27
28	`scripts/run.py` should:
29	- parse the extraction schema from the protocol
30	- filter `include` rows
31	- materialize a normalized CSV
32
33	## Acceptance
34
35	- output exists
36	- include rows map 1:1 to extraction rows
37	- schema matches `output/PROTOCOL.md`
38
39	## Non-goals
40
41	- synthesis writing
42	- bias scoring

People who liked this also viewed

6 picks

Type

extraction-form

489stars39forks

View on GitHub

How to install extraction-form?

willoscar/research-units-pipeline-skills/extraction-form

$npx -y skills add WILLOSCAR/research-units-pipeline-skills --skill extraction-form

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Extraction Form
2
3	Transforms screened include rows plus protocol schema into the analysis table used by `evidence-review`.
4
5	## Inputs
6
7	Required:
8	- `papers/screening_log.csv`
9	- `output/PROTOCOL.md`
10
11	Optional:
12	- `papers/paper_notes.jsonl`
13
14	## Output
15
16	- `papers/extraction_table.csv`
17
18	## Contract
19
20	The table must:
21	- contain one row per included paper
22	- preserve provenance columns (`paper_id`, `title`, `year`, `url`)
23	- include protocol-defined extraction fields
24	- keep narrative residue in `notes`, not in schema columns
25
26	## Script boundary
27
28	`scripts/run.py` should:
29	- parse the extraction schema from the protocol
30	- filter `include` rows
31	- materialize a normalized CSV
32
33	## Acceptance
34
35	- output exists
36	- include rows map 1:1 to extraction rows
37	- schema matches `output/PROTOCOL.md`
38
39	## Non-goals
40
41	- synthesis writing
42	- bias scoring

People who liked this also viewed

6 picks

Type