policyengine-research-lookup

Name: policyengine-research-lookup
Author: PolicyEngine

31stars5forks

How to install policyengine-research-lookup?

policyengine/policyengine-claude/policyengine-research-lookup

$npx -y skills add PolicyEngine/policyengine-claude --skill policyengine-research-lookup

Installs into the current project.

Summary

Find and cite PolicyEngine's published blog posts and research articles as evidence or proof points — for talks, pitches, grant narratives, and "has PolicyEngine written about X?" Triggers: "find a blog post", "PolicyEngine research on", "published analysis", "proof point", "has

Files · 1

View on GitHub

SKILL.md

1	# PolicyEngine research lookup
2
3	Find existing PolicyEngine articles to cite as evidence or to model tone against. PolicyEngine
4	has published ~185 articles — US and UK policy analyses, model/methodology notes, and org
5	announcements. Paths below are verified against the policyengine-app-v2 mirror.
6
7	## Where posts live
8
9	In policyengine-app-v2:
10
11	- Article bodies (plain markdown, some Jupyter `.ipynb`): `app/src/data/posts/articles/<filename>`
12	- Post index (title, description, date, tags, authors, image): `app/src/data/posts/posts.json`
13	- Author records (slug → name/bio): `app/src/data/posts/authors.json`
14
15	Article files are plain markdown — metadata is not in per-file YAML frontmatter. Each post's
16	title, date, tags, and authors live in `posts.json`, keyed to the article's `filename`; the
17	`authors` field holds slugs that resolve in `authors.json`. A `posts.json` entry:
18
19	```json
20	{
21	"title": "How we used Claude Code to apply for the Public Benefit Innovation Fund",
22	"description": "Treating writing as software development...",
23	"date": "2025-08-28",
24	"tags": ["us", "org", "ai", "featured"],
25	"authors": ["max-ghenis"],
26	"filename": "policyengine-atlas-pbif-grant.md",
27	"image": "policyengine-atlas-demo.png"
28	}
29	```
30
31	A post's public URL is `policyengine.org/<country>/research/<filename-without-extension>`.
32
33	## Lookup recipes
34
35	```bash
36	POSTS=~/PolicyEngine/policyengine-app-v2/app/src/data/posts
37
38	# Full-text search article bodies for a topic
39	grep -rli "child tax credit" "$POSTS/articles/" \| head
40
41	# Search titles/descriptions in the index
42	python3 -c 'import json;[print(p["date"],p["filename"]) for p in json.load(open("'"$POSTS"'/posts.json")) if "eitc" in (p["title"]+p.get("description","")).lower()]'
43
44	# Filter by tag (e.g. everything tagged "featured" or "uk")
45	python3 -c 'import json;[print(p["date"],p["title"]) for p in json.load(open("'"$POSTS"'/posts.json")) if "featured" in p.get("tags",[])]'
46
47	# Most recent posts
48	python3 -c 'import json;[print(p["date"],p["title"]) for p in sorted(json.load(open("'"$POSTS"'/posts.json")),key=lambda p:p["date"],reverse=True)[:10]]'
49	```
50
51	Prefer these recipes over any hardcoded list — the corpus changes weekly, so a static index of
52	filenames goes stale. Search live, then cite the exact filename and date you found.
53
54	## Frequently cited proof points
55
56	Verified filenames worth knowing (confirm current framing in the index before quoting):
57
58	\| Topic \| filename \| Use for \|
59	\|---\|---\|---\|
60	\| UK government adoption \| `policyengine-10-downing-street.md` \| credibility, state capacity \|
61	\| US state tax coverage \| `state-tax-model-beta.md` \| federal + 50-state reach \|
62	\| Grant-writing with AI \| `policyengine-atlas-pbif-grant.md` \| tooling, org velocity \|
63	\| US federal proposals \| `harris-ctc.md`, `harris-eitc.md` \| CTC/EITC analysis depth \|
64	\| UK fiscal events \| `autumn-budget-2024-employer-nic-pension-contributions.md` \| budget/manifesto costing \|
65
66	## Related skills
67
68	- policyengine-writing — match the tone and neutrality standard of published posts.
69	- policyengine-us / policyengine-uk — for "do we cover program/state X?" answer from the
70	model, not just the blog.

People who liked this also viewed

6 picks

Type

policyengine-research-lookup

31stars5forks

View on GitHub

How to install policyengine-research-lookup?

policyengine/policyengine-claude/policyengine-research-lookup

$npx -y skills add PolicyEngine/policyengine-claude --skill policyengine-research-lookup

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# PolicyEngine research lookup
2
3	Find existing PolicyEngine articles to cite as evidence or to model tone against. PolicyEngine
4	has published ~185 articles — US and UK policy analyses, model/methodology notes, and org
5	announcements. Paths below are verified against the policyengine-app-v2 mirror.
6
7	## Where posts live
8
9	In policyengine-app-v2:
10
11	- Article bodies (plain markdown, some Jupyter `.ipynb`): `app/src/data/posts/articles/<filename>`
12	- Post index (title, description, date, tags, authors, image): `app/src/data/posts/posts.json`
13	- Author records (slug → name/bio): `app/src/data/posts/authors.json`
14
15	Article files are plain markdown — metadata is not in per-file YAML frontmatter. Each post's
16	title, date, tags, and authors live in `posts.json`, keyed to the article's `filename`; the
17	`authors` field holds slugs that resolve in `authors.json`. A `posts.json` entry:
18
19	```json
20	{
21	"title": "How we used Claude Code to apply for the Public Benefit Innovation Fund",
22	"description": "Treating writing as software development...",
23	"date": "2025-08-28",
24	"tags": ["us", "org", "ai", "featured"],
25	"authors": ["max-ghenis"],
26	"filename": "policyengine-atlas-pbif-grant.md",
27	"image": "policyengine-atlas-demo.png"
28	}
29	```
30
31	A post's public URL is `policyengine.org/<country>/research/<filename-without-extension>`.
32
33	## Lookup recipes
34
35	```bash
36	POSTS=~/PolicyEngine/policyengine-app-v2/app/src/data/posts
37
38	# Full-text search article bodies for a topic
39	grep -rli "child tax credit" "$POSTS/articles/" \| head
40
41	# Search titles/descriptions in the index
42	python3 -c 'import json;[print(p["date"],p["filename"]) for p in json.load(open("'"$POSTS"'/posts.json")) if "eitc" in (p["title"]+p.get("description","")).lower()]'
43
44	# Filter by tag (e.g. everything tagged "featured" or "uk")
45	python3 -c 'import json;[print(p["date"],p["title"]) for p in json.load(open("'"$POSTS"'/posts.json")) if "featured" in p.get("tags",[])]'
46
47	# Most recent posts
48	python3 -c 'import json;[print(p["date"],p["title"]) for p in sorted(json.load(open("'"$POSTS"'/posts.json")),key=lambda p:p["date"],reverse=True)[:10]]'
49	```
50
51	Prefer these recipes over any hardcoded list — the corpus changes weekly, so a static index of
52	filenames goes stale. Search live, then cite the exact filename and date you found.
53
54	## Frequently cited proof points
55
56	Verified filenames worth knowing (confirm current framing in the index before quoting):
57
58	\| Topic \| filename \| Use for \|
59	\|---\|---\|---\|
60	\| UK government adoption \| `policyengine-10-downing-street.md` \| credibility, state capacity \|
61	\| US state tax coverage \| `state-tax-model-beta.md` \| federal + 50-state reach \|
62	\| Grant-writing with AI \| `policyengine-atlas-pbif-grant.md` \| tooling, org velocity \|
63	\| US federal proposals \| `harris-ctc.md`, `harris-eitc.md` \| CTC/EITC analysis depth \|
64	\| UK fiscal events \| `autumn-budget-2024-employer-nic-pension-contributions.md` \| budget/manifesto costing \|
65
66	## Related skills
67
68	- policyengine-writing — match the tone and neutrality standard of published posts.
69	- policyengine-us / policyengine-uk — for "do we cover program/state X?" answer from the
70	model, not just the blog.

People who liked this also viewed

6 picks

Type