research

Name: research
Author: Borda

23stars3forks

How to install research?

borda/ai-rig/research

$npx -y skills add Borda/AI-Rig --skill research

Installs into the current project.

Summary

Minimal codex-native research loop. Use for docs/papers/state-of-the-art scan with source-backed recommendations.

Files · 1

View on GitHub

SKILL.md

1	# Research
2
3	Source-backed research for documentation, API migration, paper, or state-of-the-art questions.
4
5	## Input Schema
6
7	```json
8	{
9	"question": "required research question",
10	"mode": "docs\|sota\|paper\|methodology\|code-fidelity",
11	"constraints": [
12	"optional codebase, compute, version, or implementation constraints"
13	],
14	"done_when": "recommendations are source-backed with caveats and confidence"
15	}
16	```
17
18	## Workflow
19
20	### 01: Create run directory
21
22	Run `python PLUGIN_ROOT/shared/create_run.py --skill research` once. Retain its single printed path as
23	`<run-directory>` and substitute that literal path into every later artifact path and helper argument. Never store or
24	reuse the path through a shell variable; shell variables do not persist across tool calls.
25
26	### 02: Define research question, mode, and constraints
27
28	Modes:
29
30	- `docs`: current API/docs/migration answer.
31	- `sota`: method comparison and implementation recommendation.
32	- `paper`: single-paper analysis.
33	- `methodology`: experiment design, metric, guard, and ablation review.
34	- `code-fidelity`: compare paper/spec claims against implementation.
35
36	### 03: Gather sources
37
38	Write `<run-directory>/sources.md`:
39
40	```markdown
41	\| Source \| Type \| Date/version \| Why reliable \| Used for \|
42	\| --- \| --- \| --- \| --- \| --- \|
43	```
44
45	Source rules:
46
47	- Prefer primary docs, papers, specs, release notes, code.
48	- Use current live sources for volatile docs, dependencies, APIs.
49	- Mark stale/unavailable source explicitly.
50	- Do not cite secondary summaries for high-impact claims unless independently corroborated.
51
52	For `sota`, `paper`, `methodology`, `code-fidelity`, apply `../../shared/specialist-orchestration.md` when independent expertise improves correctness. Write `<run-directory>/specialist-research-plan.md` with context packs for:
53
54	- `web-explorer`: current docs, release notes, API and dependency changes.
55	- `scientist`: formulas, methodology, metrics, ablations, benchmark claims.
56	- `solution-architect`: implementation fit, API boundaries, migration shape.
57	- `squeezer`: performance or resource claims.
58	- `data-steward`: datasets, splits, leakage, reproducibility.
59	- `challenger`: unsupported recommendation or overconfident source synthesis.
60
61	Do not send full papers, repositories, or all search results to every specialist. Give each only source excerpts, code files, questions needed for its axis.
62
63	### 04: Map to codebase context when implementation is relevant
64
65	Inspect `python PLUGIN_ROOT/shared/collect_diff.py --help`; collect `working-tree` scope into `<run-directory>/baseline`. Run topic scan separately; record unavailable paths/collection failures as evidence gaps.
66
67	### 05: Produce `<run-directory>/research.md` with:
68
69	- `Question`
70	- `Constraints`
71	- `Source Table`
72	- `Findings`
73	- `Comparison` for SOTA/method choices
74	- `Implementation Fit`
75	- `Risks And Unknowns`
76	- `Recommendation`
77	- `Next Checks`
78
79	### 06: For paper/code-fidelity mode, include dimensions:
80
81	- `F`: formula/math match
82	- `H`: hyperparameter parity
83	- `E`: evaluation protocol
84	- `N`: notation and naming consistency
85	- `C`: citation/derivation chain
86
87	### 07: Run review gate
88
89	Run `git diff --check` as an argv command. Write its combined output to `<run-directory>/review.txt` and retain its
90	exit status as review evidence; do not erase a nonzero result.
91
92	### 08: Run shared gates and write the validated result artifact
93
94	Follow `../../shared/helper-cli-contract.md` and authoritative help. For research-only work, mark lint, format, types, tests inapplicable with concrete reasons; review requires non-empty `research.md`, `sources.md`, clean diff check. Write `RESEARCH_METADATA`, validate as `research`, promote only validated candidate.
95
96	Replace explicit skip with relevant command when research includes executable validation.
97
98	## Fail-Fast Rules
99
100	1. Missing question => fail.
101	2. No primary sources for high-impact/current claims => fail.
102	3. Recommendation not tied to constraints => fail.
103	4. Paper/code-fidelity claim without code or source reference => fail.
104	5. Result artifact missing => fail.
105
106	## Quality Gates
107
108	Required:
109
110	- `review`: source table, caveats, self-review, `git diff --check`.
111
112	Conditional:
113
114	- `tests`: when research includes an executable validation or code-fidelity probe.
115
116	## Calibration Hooks
117
118	On source-protocol/recommendation-policy change, update calibration:
119
120	- behavioral cases: stale docs, unsupported SOTA claim, paper-code mismatch
121	- benchmark patterns: `research`
122
123	## Output Contract
124
125	Use shared gate schema from `../../shared/quality-gates.md`.
126
127	Minimum artifact payload template: `result-template.json`.

People who liked this also viewed

6 picks

Type

azure-kustoQuery and analyze data in Azure Data Explorer (Kusto/ADX) using KQL for log analytics, telemetry, and time series analysis. WHEN: KQL queries, Kusto database queries, Azure Data Explorer, ADX clustersSkillsJul 2026470k1.3k
lark-base飞书多维表格（Base）操作：建表、字段、记录、视图、统计、公式/lookup、表单、仪表盘、workflow、角色权限；遇到 Base/多维表格/bitable 或 /base/ 链接时使用。文件导入转 lark-drive，认证/授权转 lark-shared。SkillsJul 2026382k16k
lark-calendar飞书日历：管理日历日程和会议室。查看/搜索日程、创建/更新日程、管理参会人、查询忙闲和推荐时段、预定会议室。当用户需要查看日程安排、创建/修改会议、查询/预定会议室时使用。不负责：查询过去的视频会议记录（走 lark-vc）、待办任务（走 lark-task）。SkillsJul 2026380k16k
shadcnManages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. AppliSkillsJul 2026246k120k
xlsxUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.SkillsJul 2026139k163k
firebase-data-connectAgent Skills for FirebaseSkillsJul 2026108k387

research

23stars3forks

View on GitHub

How to install research?

borda/ai-rig/research

$npx -y skills add Borda/AI-Rig --skill research

Installs into the current project.

Summary

Minimal codex-native research loop. Use for docs/papers/state-of-the-art scan with source-backed recommendations.

Files · 1

View on GitHub

SKILL.md

1	# Research
2
3	Source-backed research for documentation, API migration, paper, or state-of-the-art questions.
4
5	## Input Schema
6
7	```json
8	{
9	"question": "required research question",
10	"mode": "docs\|sota\|paper\|methodology\|code-fidelity",
11	"constraints": [
12	"optional codebase, compute, version, or implementation constraints"
13	],
14	"done_when": "recommendations are source-backed with caveats and confidence"
15	}
16	```
17
18	## Workflow
19
20	### 01: Create run directory
21
22	Run `python PLUGIN_ROOT/shared/create_run.py --skill research` once. Retain its single printed path as
23	`<run-directory>` and substitute that literal path into every later artifact path and helper argument. Never store or
24	reuse the path through a shell variable; shell variables do not persist across tool calls.
25
26	### 02: Define research question, mode, and constraints
27
28	Modes:
29
30	- `docs`: current API/docs/migration answer.
31	- `sota`: method comparison and implementation recommendation.
32	- `paper`: single-paper analysis.
33	- `methodology`: experiment design, metric, guard, and ablation review.
34	- `code-fidelity`: compare paper/spec claims against implementation.
35
36	### 03: Gather sources
37
38	Write `<run-directory>/sources.md`:
39
40	```markdown
41	\| Source \| Type \| Date/version \| Why reliable \| Used for \|
42	\| --- \| --- \| --- \| --- \| --- \|
43	```
44
45	Source rules:
46
47	- Prefer primary docs, papers, specs, release notes, code.
48	- Use current live sources for volatile docs, dependencies, APIs.
49	- Mark stale/unavailable source explicitly.
50	- Do not cite secondary summaries for high-impact claims unless independently corroborated.
51
52	For `sota`, `paper`, `methodology`, `code-fidelity`, apply `../../shared/specialist-orchestration.md` when independent expertise improves correctness. Write `<run-directory>/specialist-research-plan.md` with context packs for:
53
54	- `web-explorer`: current docs, release notes, API and dependency changes.
55	- `scientist`: formulas, methodology, metrics, ablations, benchmark claims.
56	- `solution-architect`: implementation fit, API boundaries, migration shape.
57	- `squeezer`: performance or resource claims.
58	- `data-steward`: datasets, splits, leakage, reproducibility.
59	- `challenger`: unsupported recommendation or overconfident source synthesis.
60
61	Do not send full papers, repositories, or all search results to every specialist. Give each only source excerpts, code files, questions needed for its axis.
62
63	### 04: Map to codebase context when implementation is relevant
64
65	Inspect `python PLUGIN_ROOT/shared/collect_diff.py --help`; collect `working-tree` scope into `<run-directory>/baseline`. Run topic scan separately; record unavailable paths/collection failures as evidence gaps.
66
67	### 05: Produce `<run-directory>/research.md` with:
68
69	- `Question`
70	- `Constraints`
71	- `Source Table`
72	- `Findings`
73	- `Comparison` for SOTA/method choices
74	- `Implementation Fit`
75	- `Risks And Unknowns`
76	- `Recommendation`
77	- `Next Checks`
78
79	### 06: For paper/code-fidelity mode, include dimensions:
80
81	- `F`: formula/math match
82	- `H`: hyperparameter parity
83	- `E`: evaluation protocol
84	- `N`: notation and naming consistency
85	- `C`: citation/derivation chain
86
87	### 07: Run review gate
88
89	Run `git diff --check` as an argv command. Write its combined output to `<run-directory>/review.txt` and retain its
90	exit status as review evidence; do not erase a nonzero result.
91
92	### 08: Run shared gates and write the validated result artifact
93
94	Follow `../../shared/helper-cli-contract.md` and authoritative help. For research-only work, mark lint, format, types, tests inapplicable with concrete reasons; review requires non-empty `research.md`, `sources.md`, clean diff check. Write `RESEARCH_METADATA`, validate as `research`, promote only validated candidate.
95
96	Replace explicit skip with relevant command when research includes executable validation.
97
98	## Fail-Fast Rules
99
100	1. Missing question => fail.
101	2. No primary sources for high-impact/current claims => fail.
102	3. Recommendation not tied to constraints => fail.
103	4. Paper/code-fidelity claim without code or source reference => fail.
104	5. Result artifact missing => fail.
105
106	## Quality Gates
107
108	Required:
109
110	- `review`: source table, caveats, self-review, `git diff --check`.
111
112	Conditional:
113
114	- `tests`: when research includes an executable validation or code-fidelity probe.
115
116	## Calibration Hooks
117
118	On source-protocol/recommendation-policy change, update calibration:
119
120	- behavioral cases: stale docs, unsupported SOTA claim, paper-code mismatch
121	- benchmark patterns: `research`
122
123	## Output Contract
124
125	Use shared gate schema from `../../shared/quality-gates.md`.
126
127	Minimum artifact payload template: `result-template.json`.

People who liked this also viewed

6 picks

Type