research
v0.2.28by Vycari · author: Allen Hutchison
Proactive research skills that surface relevant outside knowledge and open a PR or digest. research-radar periodically scans arXiv for papers relevant to the work in this repo and writes a dated digest. A home for future research tooling (deep research, GitHub/code search, prior-art scans). Themes and output paths come from.claude/maintainerd.json.
What research does, in the publisher's words
Proactive reading, scoped to what this repo is actually building. The value is entirely in the scoping, a generic "AI papers this week" digest is worthless.
arXiv is an open submission channel, so paper titles and abstracts are text a stranger wrote, and the digest this skill produces is committed to your repo. It summarizes rather than reproduces, and follows references/untrusted-input.md for anything addressed to automation. It also reports only what the abstract claims, it reads abstracts, not full papers, and says so.
From the project README, punctuation lightly normalized · Open the README on GitHub
- Skills
- 1
- MCP servers
- 0
- Stars
- 3
- License
- MIT
- Repo created
- 2026-06-28
- Last pushed
- 2026-10-08
- Publisher type
- Organization
- Version
- 0.2.28
Links
Skills · 1
- research-radar
- Weekly scan of arXiv for papers relevant to the work in this repo, derive themes from the repo, query arXiv, curate the few most relevant papers, write a dated digest, and open a PR. Use when the user asks to "run the research radar", "scan arxiv", "find recent papers relevant to <repo>", "what sh…
Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.
Category
Research & Search. Assigned by this directory. Agent Plugins 1.0.0 has no category field, so no manifest declares one.
Keywords
research · arxiv · digest · discovery · maintainer
Related by publisher and keywords
Scheduled findings that open PRs/issues, deliberately capped and conservative: audit-architecture (tech-debt sweep), audit-tests (test-suite health), audit-security (vulnerable deps, committed secrets incl. git history, permissive defaults: code patterns are left to Claude Code's built-in /security-review), audit-deps (dependency health: outdated, deprecated, unused, lockfile drift, licenses), audit-design-docs (validate design docs against code), and audit-product-docs (validate user-facing docs against code). Detection rules and invariants come from the repo's.claude/maintainerd.json and guidelines files.
Autonomous issue-to-PR pipeline. create-issue is the front door: it turns a rough request into a well-formed, buildable issue by investigating the codebase. auto-dev runs one tick of triage -> plan -> build -> address-review against an auto:* GitHub label state machine and never merges; review-queue is the maintainer's console for approving plans, answering questions, and merging. Label names, comment markers, and branch prefixes are configurable via.claude/maintainerd.json so the pipeline drops into any repo.
Autonomous Dependabot queue management: the one maintainerd plugin that merges. The dependabot skill gates every bot-authored dependency PR on all checks green, no requested changes, and a configured semver policy (patch/minor by default, majors held for the human), merges one PR per non-overlapping file group per pass, waits for Dependabot to rebase the rest (nudging with @dependabot rebase only when a rebase is genuinely stuck), and repeats until the queue drains. It never fixes a broken update: it diagnoses the CI failure, files one issue with the evidence, labels the PR blocked, and moves on. Install only in repos where a bot is allowed to merge; every knob lives in.claude/maintainerd.json's depsFlow block.
User-scoped narration and record-keeping that spans repos. worklog captures a day's shipped work, from merged PRs enriched by the live session, into the user's Obsidian vault as a session-summary note, a project-hub link, and a daily-note line. Unlike the repo-scoped plugins, its vault setting lives in a user-level ~/.claude/maintainerd.json, with an optional per-repo project pointer. A home for future journal skills (weekly-review, decision-log).
Foundation for the maintainerd maintainer toolkit. Ships bootstrap (generates a repo's.claude/maintainerd.json config contract and starter guidelines), doctor (validates that setup, config, paths, commands, labels, guidelines, rosters, the coverage ratchet, and, with --profile, reports a repo's drift from a fleet's repo standard), and new-repo (creates or adopts a repo against that standard), plus the canonical config and profile schema references that every other maintainerd plugin reads.
Baseline maintainer dev flow as config-driven skills: create-pr (opens PRs only after running the repo's format/lint/build/test pre-flight), address-review (drive the iterative response loop on bot + human PR feedback to approval, fix, push, reply to every inline thread and a PR-level summary), release (cut a versioned release: gather changes since the last tag, update notes, run the gate, bump, tag, publish), daily-changelog (one markdown file per day from merged PRs), and daily-update (runs the repo's per-day housekeeping skills and bundles their output into one PR). Also ships three PreToolUse Bash hooks: pr-template-guard (denies a gh pr create/edit whose body is missing a heading the repo's own PR template requires), skip-label-race-guard (warns when gh pr create applies the repo's review-skip label without --draft), merge-guard (warns, never denies, when a command merges a pull request), plus an opt-in PostToolUse hook, review-reply-postcondition (a review reply that claims a fix must name a commit reachable from HEAD). Reads.claude/maintainerd.json.
[](https://agentpluginsdirectory.com/plugins/research--vycari)✓ Verified . We fetched the manifest from GitHub and checked it against the official Agent Plugins 1.0.0 schema at agent-plugins.org.