/deep-research — Multi-source research with archive
Plans, executes, and synthesizes research from multiple sources. Archives the output so the corpus compounds.
Step 1 — Frame the question
Restate the research question in one tight sentence. If ambiguous, ask the user:
- What's the decision this research will inform?
- What's the minimum useful answer? (Saves over-researching.)
- Any sources to prioritize or avoid?
Output: **Research question:** <one sentence>
Step 2 — Plan the sources
Pick from this menu based on the question type. Note which sources you'll hit and why.
| Source |
When to use |
Tool |
| Web search (Google) |
Authoritative articles, docs, official statements |
WebSearch |
/last30days |
What people are actually saying right now — Reddit, X, YouTube, HN, web recency |
Skill({skill: "last30days", args: "<topic>"}) |
| Specific URLs |
When the user hands over starting URLs |
WebFetch |
| Browsable pages (auth-walled, JS-heavy) |
Pricing pages, product tours, profiles |
agent-browser via the compound-engineering:agent-browser skill |
| Memory |
Prior research / decisions / context the user already captured |
grep ~/.claude/memory/ |
| Notion |
If the topic touches a known Notion workspace |
Direct Notion API (key in $NOTION_API_KEY, see reference_notion_api.md) |
| Research archive |
Prior /deep-research runs that touched this topic |
grep ${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/deep-research/archive/ |
Run discovery passes in parallel where possible. Sequential only when one source needs another's output (e.g., agent-browser a URL discovered by WebSearch).
Step 3 — Execute discovery
Run each chosen source. For each result, capture:
- The source (URL or system)
- 1–3 sentence summary of what was said
- Date / recency
- Confidence in the source (high/medium/low)
Don't synthesize yet — just collect.
Step 4 — Synthesize
- Group findings by theme or sub-question
- Contradiction check — flag anywhere sources disagree. Don't average them; surface the disagreement.
- Confidence: high (multiple independent sources agree), medium (one strong source or several weak), low (single anecdote or speculation)
- Gaps: what would change the answer? What's NOT in the corpus?
Step 5 — Output the brief
Use this template:
# Research: <question>
**Date:** <YYYY-MM-DD>
**Decision this informs:** <one line>
**Confidence overall:** high / medium / low
## TL;DR
<2–4 sentences with the answer>
## Key findings
### 1. <Finding>
<2–4 sentences>. Sources: [1], [3], [5]
### 2. <Finding>
...
## Contradictions / uncertainty
- <where sources disagree, with each side cited>
## Gaps
- <what's missing from the corpus>
- <what to research next to close the gap>
## Recommended next steps
1. <action>
2. <action>
## Sources
[1] <Title> — <URL or system> (<date>) — <confidence>
[2] ...
Step 6 — Archive
Archives live in ${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/deep-research/archive/ (create the directory if missing). Never write archives inside the skill's own folder — skill installs and upgrades re-sync from source and wipe anything saved there. Migration: if this skill's folder contains an old references/research-archive/ with user entries, move those files into the archive directory first.
Write the brief to <archive dir>/<YYYY-MM-DD>-<slug>.md so it's grep-able forever. Slug = kebab-case of the topic.
Also append a one-line entry to <archive dir>/INDEX.md (create if missing):
- 2026-06-15 — [<topic>](./<filename>.md) — <one-line TL;DR>
Step 7 — Surface
After archiving:
- Show the full brief in chat
- Tell the user the archive path
- Offer: "Push to Notion or save to a project's docs?"
Composes with
business-brainstorm — calls this skill during the market validation step/domain — when research includes "is the .com available"/last30days — one of the data sources
Notes on quality
- Always cite. Every claim in the brief needs a source pointer.
- Recency matters — note dates on each source. For fast-moving topics (AI, startups), de-weight sources >12 months old.
- Don't trust a single source for high-stakes claims. Re-search until you have at least 2 independent corroborations or surface the uncertainty.
- No padding. If the answer is one paragraph, return one paragraph. The template is a maximum, not a minimum.
1---2name: deep-research-23description: When you want multi-source, multi-step research on a topic — competitor research before a sales call, market research for a new business idea, positioning angles, due diligence on a partnership or podcast guest, tech decision research (which DB, which auth), or any "I need to actually understand X." Combines WebSearch, WebFetch, agent-browser, /last30days (Reddit/X/YouTube/HN/web recency), memory, and Notion. Outputs a structured brief with citations, contradictions, gaps, and recommended next steps. Archives every research run to ~/.config/makerskills/deep-research/archive/ so past work is searchable. Triggers on "/deep-research," "research X," "investigate X," "do a deep dive on X," "look into X," "what's actually happening with X," "due diligence on X," "validate this market." Differs from a one-shot WebSearch: this is multi-pass with verification.4---5
6# /deep-research — Multi-source research with archive
7
8Plans, executes, and synthesizes research from multiple sources. Archives the output so the corpus compounds.
9
10## Step 1 — Frame the question
11
12Restate the research question in one tight sentence. If ambiguous, ask the user:
13- What's the decision this research will inform?
14- What's the minimum useful answer? (Saves over-researching.)
15- Any sources to prioritize or avoid?
16
17Output: `**Research question:** <one sentence>`
18
19## Step 2 — Plan the sources
20
21Pick from this menu based on the question type. Note which sources you'll hit and why.
22
23| Source | When to use | Tool |
24|---|---|---|
25| Web search (Google) | Authoritative articles, docs, official statements | `WebSearch` |
26| `/last30days` | What people are *actually saying* right now — Reddit, X, YouTube, HN, web recency | `Skill({skill: "last30days", args: "<topic>"})` |
27| Specific URLs | When the user hands over starting URLs | `WebFetch` |
28| Browsable pages (auth-walled, JS-heavy) | Pricing pages, product tours, profiles | `agent-browser` via the `compound-engineering:agent-browser` skill |
29| Memory | Prior research / decisions / context the user already captured | grep `~/.claude/memory/` |
30| Notion | If the topic touches a known Notion workspace | Direct Notion API (key in `$NOTION_API_KEY`, see `reference_notion_api.md`) |
31| Research archive | Prior `/deep-research` runs that touched this topic | grep `${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/deep-research/archive/` |
32
33Run discovery passes **in parallel** where possible. Sequential only when one source needs another's output (e.g., agent-browser a URL discovered by WebSearch).
34
35## Step 3 — Execute discovery
36
37Run each chosen source. For each result, capture:
38- The source (URL or system)
39- 1–3 sentence summary of what was said
40- Date / recency
41- Confidence in the source (high/medium/low)
42
43Don't synthesize yet — just collect.
44
45## Step 4 — Synthesize
46
471. **Group findings** by theme or sub-question
482. **Contradiction check** — flag anywhere sources disagree. Don't average them; surface the disagreement.
493. **Confidence**: high (multiple independent sources agree), medium (one strong source or several weak), low (single anecdote or speculation)
504. **Gaps**: what would change the answer? What's NOT in the corpus?
51
52## Step 5 — Output the brief
53
54Use this template:
55
56```markdown
57# Research: <question>
58
59**Date:** <YYYY-MM-DD>
60**Decision this informs:** <one line>
61**Confidence overall:** high / medium / low
62
63## TL;DR
64<2–4 sentences with the answer>
65
66## Key findings
67
68### 1. <Finding>
69<2–4 sentences>. Sources: [1], [3], [5]
70
71### 2. <Finding>
72...
73
74## Contradictions / uncertainty
75- <where sources disagree, with each side cited>
76
77## Gaps
78- <what's missing from the corpus>
79- <what to research next to close the gap>
80
81## Recommended next steps
821. <action>
832. <action>
84
85## Sources
86[1] <Title> — <URL or system> (<date>) — <confidence>
87[2] ...
88```
89
90## Step 6 — Archive
91
92Archives live in `${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/deep-research/archive/` (create the directory if missing). Never write archives inside the skill's own folder — skill installs and upgrades re-sync from source and wipe anything saved there. **Migration:** if this skill's folder contains an old `references/research-archive/` with user entries, move those files into the archive directory first.
93
94Write the brief to `<archive dir>/<YYYY-MM-DD>-<slug>.md` so it's grep-able forever. Slug = kebab-case of the topic.
95
96Also append a one-line entry to `<archive dir>/INDEX.md` (create if missing):
97
98```markdown
99- 2026-06-15 — [<topic>](./<filename>.md) — <one-line TL;DR>
100```
101
102## Step 7 — Surface
103
104After archiving:
105- Show the full brief in chat
106- Tell the user the archive path
107- Offer: *"Push to Notion or save to a project's docs?"*
108
109## Composes with
110
111- `business-brainstorm` — calls this skill during the market validation step
112- `/domain` — when research includes "is the .com available"
113- `/last30days` — one of the data sources
114
115## Notes on quality
116
117- **Always cite.** Every claim in the brief needs a source pointer.
118- **Recency matters** — note dates on each source. For fast-moving topics (AI, startups), de-weight sources >12 months old.
119- **Don't trust a single source** for high-stakes claims. Re-search until you have at least 2 independent corroborations or surface the uncertainty.
120- **No padding.** If the answer is one paragraph, return one paragraph. The template is a maximum, not a minimum.
majiayu000