一键在 Manus 中运行任何 Skill

autoresearch-db

星标58

分支9

更新时间2026年7月19日 03:15

Use when building the evidence base for a chosen autoresearch topic — maps the insight areas needed to propose new ideas, downloads references to cover them via download-ref, distills each area into research/INSIGHTS.md for user selection, builds a structured domain database and pinned reference implementations, and catalogs all algorithms/software in research/CATALOG.md. Owns the survey gate. Stage 2 of the autoresearch pipeline.

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

在 Manus 中运行

来源

QuantumBFS

QuantumBFS/sci-brain

打开 GitHub 仓库查看创作者相关仓库

下载

在 Manus 中运行

Autoresearch DB

Input: a topic with metrics in topics.md. Output: a knowledge base, an insight file, a domain database, pinned reference code, a catalog — and the survey gate passed. The point is not "papers exist on disk" but "the agent holds the insights needed to propose new ideas".

Procedure

Map insight areas. From the topic and its metrics, list the insight areas an idea-proposer needs: algorithmic techniques, proof/analysis methods, data structures, benchmark practices for this problem class. Show the list to the user before downloading anything.
Download for coverage. Acquire references with the download-ref skill into <project>/.knowledge/. Coverage check: every insight area is covered by ≥1 downloaded reference; areas with no coverage trigger further search/downloads. SOTA-results-only coverage is insufficient.
Distill. For each insight area, write one entry in research/INSIGHTS.md following skills/autoresearch-db/references/insights-template.md (Technique / Applies when / Limits / Sources). Distill from the rendered papers in .knowledge/, not from memory.
User selects. Present the distilled areas via AskUserQuestion (multi-select): which are needed for idea generation on this topic? Selected entries go under ## Selected, the rest under ## Shelved. The run loop draws only on Selected.
Domain database. Build the structured dataset the topic needs (e.g. QEC codes as JSON) under research/database/, with a README.md documenting schema and provenance of every record.
Pin reference implementations. For each key algorithm with public code: record URL + commit hash, build it, and run its own smoke test/example. Implement small reference versions only where no code exists and the algorithm is load-bearing.
Catalog. Write research/CATALOG.md: every algorithm/software relevant to the topic, one row each — name, source (paper/repo), status reproduced (we ran it) / pinned (source-locked, not yet run) / paper-only (no code exists), and notes.

Survey gate (owned by this skill)

Flip survey_gate: passed <date> and set stage: validator in research/STATE.md only when all of:

research/CATALOG.md exists and every key algorithm from the insight map appears in it;
.knowledge/INDEX.md lists every reference cited in INSIGHTS.md and CATALOG.md;
research/INSIGHTS.md has a non-empty ## Selected section chosen by the user;
every reproduced-status entry actually ran (its output is recorded in CATALOG.md notes).

同仓库更多 Skills

同仓库

autoresearch-run

QuantumBFS/sci-brain

Use when running the autoresearch loop for a project whose survey and validator gates have passed — executes batches of attempts (one git worktree + LOG.md each, scored only by the validator under a hard time limit), then reflects, reports, and re-plans; continues autonomously while authorized_rounds remain, otherwise stops for user review. Stage 4 of the autoresearch pipeline.

2026-07-1958

autoresearch-validator

QuantumBFS/sci-brain

Use when setting the goal and building the validator for an autoresearch topic — defines the publishable bar with the user, splits dev instances from a sealed gitignored holdout, builds a Docker-canonical strict environment (recorded fallback allowed), implements the validate CLI with rich JSON errors, and passes the strictness self-test against negative controls. Owns the validator gate. Stage 3 of the autoresearch pipeline.

2026-07-1958

autoresearch

QuantumBFS/sci-brain

Use when starting or resuming an autoresearch project — reads research/STATE.md to find the current stage of the pipeline (topics → db → validator → run), verifies the previous stage's gate artifacts exist on disk, and routes to the matching stage skill. Triggers on "start autoresearch", "resume autoresearch", "autoresearch status", "what stage is my research project at".

2026-07-1958

autoresearch-topics

QuantumBFS/sci-brain

Use when choosing research topics within a domain that suit autoresearch and defining their score metrics — brainstorms candidates, scores each on machine-checkable success, cheap per-attempt evaluation, iteration headroom, and publishability, lets the user pick, then derives primary/guard metrics (with computation cost and gaming risks) for each chosen topic and writes everything to topics.md. Stage 1 of the autoresearch pipeline; invoked via the autoresearch dispatcher or directly with a domain.

2026-07-1958

slide-writer

QuantumBFS/sci-brain

Use when building a slide deck in Typst + Touying for a scientific talk, lecture, or project briefing — ships a browsable zoo of color themes, layout templates, and gadgets (callouts, theory boxes, CeTZ diagrams, pin annotations) that compose into projector-safe slides. Distinct from slide-writing (HTML, QudeLeap brand) and paper-writer (full manuscripts).

2026-07-0258

download-ref

QuantumBFS/sci-brain

Use when adding one or many new references (arXiv ID or DOI) to a sci-brain knowledge base — `<project>/.knowledge/` by default, or `advisors/<slug>/.knowledge/` when invoked from an advisor flow. Fetches metadata via Semantic Scholar, downloads PDFs (with SciHub fallback), renders to markdown, regenerates `INDEX.md`, and appends to the KB's `references.bib`. Handles both single refs and bulk-from-bib batches.

2026-06-2458

name

autoresearch-db

description

Autoresearch DB

Procedure

Map insight areas. From the topic and its metrics, list the insight areas an idea-proposer needs: algorithmic techniques, proof/analysis methods, data structures, benchmark practices for this problem class. Show the list to the user before downloading anything.
Download for coverage. Acquire references with the download-ref skill into <project>/.knowledge/. Coverage check: every insight area is covered by ≥1 downloaded reference; areas with no coverage trigger further search/downloads. SOTA-results-only coverage is insufficient.
Distill. For each insight area, write one entry in research/INSIGHTS.md following skills/autoresearch-db/references/insights-template.md (Technique / Applies when / Limits / Sources). Distill from the rendered papers in .knowledge/, not from memory.
User selects. Present the distilled areas via AskUserQuestion (multi-select): which are needed for idea generation on this topic? Selected entries go under ## Selected, the rest under ## Shelved. The run loop draws only on Selected.
Domain database. Build the structured dataset the topic needs (e.g. QEC codes as JSON) under research/database/, with a README.md documenting schema and provenance of every record.
Pin reference implementations. For each key algorithm with public code: record URL + commit hash, build it, and run its own smoke test/example. Implement small reference versions only where no code exists and the algorithm is load-bearing.
Catalog. Write research/CATALOG.md: every algorithm/software relevant to the topic, one row each — name, source (paper/repo), status reproduced (we ran it) / pinned (source-locked, not yet run) / paper-only (no code exists), and notes.

Survey gate (owned by this skill)

Flip survey_gate: passed <date> and set stage: validator in research/STATE.md only when all of:

research/CATALOG.md exists and every key algorithm from the insight map appears in it;
.knowledge/INDEX.md lists every reference cited in INSIGHTS.md and CATALOG.md;
research/INSIGHTS.md has a non-empty ## Selected section chosen by the user;
every reproduced-status entry actually ran (its output is recorded in CATALOG.md notes).