Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Research — gather sources and write provisional findings · inkeep | DecimalAI

researchby inkeep

Research — gather sources and write provisional findings

inkeep/open-knowledge-pack-knowledge-base-research

Source inkeep/open-knowledge ↗License GPL-3.0

Investigate a topic against preserved sources and write a provisional research article under `research/` in a Knowledge Base project (the `knowledge-base` starter pack). Read when asked to research a topic, compare options, synthesize sources, gather evidence, or extend an existing research doc. Carries the full procedure: scan existing coverage, agree a research rubric, capture every source verbatim before analyzing, write the article incrementally so a crash never loses work, cite every claim,

SkillScore40Maintained

Automatic

Research pack

Last verified 9/2/2026

What it changes

Passed 6 of 16 test cases with the skill, and 4 of 16 without it.

Without skill24%

With skill40%

Checks it passed that the model alone missed

Skips article writing (Step 5) and file validation (Step 7).
Maintains source ingestion (Step 3) and evidence citation disciplines.
Invokes audit with the target research file path to verify frontmatter and internal links.

and 2 more in the full report

4 fixed6 passed either way15 still failing0 broke

Pass-rate lift: +16 pp
Tokens: +207%
Agent turns: 0%
Response time: +13%

Model gemini-3.6-flashMeasured 9/2/2026v410 of 25 cases passed with skill · 16 comparable

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/inkeep-open-knowledge-pack-knowledge-base-research/SKILL.md https://app.decimal.ai/s/inkeep-open-knowledge-pack-knowledge-base-research/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add inkeep-open-knowledge-pack-knowledge-base-research, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/inkeep-open-knowledge-pack-knowledge-base-research/ and pins it in .decimal/skills.lock

decimalai skills add inkeep-open-knowledge-pack-knowledge-base-research

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/inkeep-open-knowledge-pack-knowledge-base-research/SKILL.md https://app.decimal.ai/s/inkeep-open-knowledge-pack-knowledge-base-research/SKILL.md

Raw file: /s/inkeep-open-knowledge-pack-knowledge-base-research/SKILL.md ↗ · machine-readable summary at /s/inkeep-open-knowledge-pack-knowledge-base-research.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 4 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 4

Prompt

You have finished drafting a new provisional research article research/sqlite-vector-search.md and ingested external sources into external-sources/. What specific validation tool call must be run in Step 7 before completing the workflow?

What was checked, without and with the skill
Checked	Without skill	With skill
Invokes audit with the target research file path to verify frontmatter and internal links.	Missed	Passed

case-13 · gemini-3.6-flash · 9/2/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md370 linessingle file

Research — gather sources and write provisional findings

This skill is pack guidance. The platform /open-knowledge skill (read/write/preview/linking/grounding rules) still governs every markdown operation — this layers the procedure on top.

Conduct evidence-driven research on a topic and produce a provisional research article under research/. Provisional, not canonical: research articles capture findings, trade-offs, and open questions at a point in time. They are promoted to canonical articles via the /consolidate-notes skill only when decisions solidify.

The content directory is the resolved content.dir — read it with config({ key: 'content.dir' }) if you don't already know it. Paths below are relative to it.

Three paths

Path A — Research article (DEFAULT): A persistent provisional article with status: draft and an inline sources: frontmatter list pointing at raw sources captured via the ingest procedure. This is the default unless the user explicitly opts out.
Path B — Direct answer: Findings delivered in conversation only. Requires explicit user request (e.g., "just tell me", "no doc needed", "quick answer").
Path C — Update existing research: Surgical additions/corrections to an existing research article. Triggered when the user references an existing research doc or says "update/refresh/extend."

Path A is the default because provisional articles compound over time; spoken answers do not.

Legacy reads: Existing articles may use status: provisional and string paths under sources:. Treat those as draft research and source resources. Do not mass-rewrite them; new writes use the OKF shapes below.

Autonomy mode

Mode	Behavior	How entered
Supervised (default)	Stop at the scoping gate for user rubric confirmation. Route coverage decisions interactively.	Default when a user drives the session.
Headless	Auto-confirm rubric after proposing it. Auto-select routing decisions. Skip interactive prompts. All other gates (scan, analysis, validation, grounding) still enforced.	Explicit "don't wait for me", "just proceed", "run headless" — or non-interactive container environments.

In headless mode, propose the rubric AND proceed immediately. Mark the Scoping task completed after proposing.

Mandatory execution order

Hard gates — do NOT skip ahead. If you find yourself about to run a WebFetch or WebSearch without completing Steps 0-2, STOP — you skipped a gate.

Step 0: Create workflow checkpoint tasks — ALWAYS the first action.
Step 1: Scan existing coverage + route — scan the content directory for prior work; classify coverage; present options before new research begins.
Step 2: Collaborative scoping — propose a research rubric. In Supervised mode, STOP and WAIT for user confirmation before any external fetch.
Step 3: Capture raw sources via ingest — preserve before analyzing.
Step 4: Read + analyze — third-party/external by default; first-party codebase only when the user explicitly requests.
Step 5: Write the research article — Path A only.
Step 6: Link aggressively + file valuable Q&A back.

Track record

Benchmark

6 / 16

test cases passed, vs 4 / 16 without

Adoption

Users: 6
Times used: 178

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v5 — 9/1/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

inkeep / open-knowledge

packages/server/assets/skills/packs/knowledge-base/research/SKILL.md

View on GitHub ↗

License GPL-3.0Imported 10/16/2026Version v4

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

Research — gather sources and write provisional findings

Three paths

Autonomy mode

Mandatory execution order

Track record

Benchmark

Adoption

SkillSafety

Source

Related skills