karlorz/dev-loop-research
skills/dev-loop-research/SKILL.md
Standalone research agent — scans repo health (CLI, skills, spec drift) and vault health (raw-to-page coverage, cross-link density, page quality, type coverage). Outputs prioritized work-item recommendations. Pass `high` for aggressive mode.
tools
Updated May 10, 2026
$ install --global
skillsauthnpx skillsauth add karlorz/agent-skills dev-loop-researchInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: May 10, 2026, 6:48 AM126.5s1 file scanned
SKILL.md
- name:
- dev-loop-research
- description:
- Standalone research agent — scans repo health (CLI, skills, spec drift) and vault health (raw-to-page coverage, cross-link density, page quality, type coverage). Outputs prioritized work-item recommendations. Pass `high` for aggressive mode.
- argument-hint:
- [high]
Dev-Loop Research Agent (Standalone)
Standalone entry point for the research agent. Enables invocation independent of the dev-loop idle path:
- One-shot:
/dev-loop-researchor/dev-loop-research high - Recurring:
/loop 1h dev-loop-research
Intensity Level
Parse arguments for high (case-insensitive). If present,
intensity = high; otherwise intensity = normal. Full threshold
definitions live in the companion prompt (see Execution below).
Project Config
Load the same way dev-loop does (.claude/dev-loop.config.md, then
CLAUDE.md introspection, then repo autodiscover). If session variables
from a prior REFRESH are already in scope ($BACKEND_CAPS,
$VAULT_TYPES, etc.), reuse them — do not re-read unchanged files.
Execution
Read the full research agent prompt from the dev-loop skill's companion
file at ~/.claude/plugins/cache/karlorz-agent-skills/dev-loop/<version>/research.md
(check the version directory) and execute exactly one research pass as
documented there. If the cache path is unavailable, fall back to
skills/dev-loop/research.md relative to the repo root.
The companion prompt is the canonical execution spec — it defines intensity thresholds, Track A/B steps, SYNTHESIZE scoring, idle fast-path, and save/exit protocol. This SKILL.md is the invocable wrapper only.
Rules
- One research pass per invocation — do not iterate.
- Delegate all logic to research.md; do not re-state its rules here.
Related Skills
karlorz/simplify
development
VerifiedTrustedCommunity
Review and simplify recently changed code for reuse, clarity, and efficiency while preserving behavior. Use when the user asks to simplify, refine, polish, clean up, or make code clearer, or after finishing a logical chunk of implementation that should be tightened before commit.
SKILL.mdUpdated Jun 3, 2026
karlorz/playwright-cli
tools
VerifiedTrustedCommunity
Use this skill when the user asks to open a browser, browse a website, scrape a page, automate Chrome, take a screenshot, fill out a form, click a button, or otherwise interact with a website. Includes a browser-worker agent (model: sonnet) for mechanical Chrome lifecycle and interaction tasks.
SKILL.mdUpdated Jun 3, 2026
karlorz/host-backup-restore
development
VerifiedTrustedCommunity
Host-level backup and restore with profile system (presets + custom YAML profiles), model-aware agents (sonnet worker for mechanical tasks), post-discovery research, and skillwiki infrastructure capture. Uses rsync with partial-dir for resumable WAN transfers. Use when backing up or restoring Caddy reverse-proxy domains, databases (postgres, mysql, redis, mongodb, sqlite), systemd services, full SSH identity/config, Tailscale state/config, and Hermes agent state on remote Linux hosts.
SKILL.mdUpdated Jun 3, 2026
karlorz/hermes-cli
tools
VerifiedTrustedCommunity
Hermes Agent CLI commands reference. Use when the user asks about hermes-agent CLI usage, commands, flags, or subcommands. Covers the full hermes terminal command surface.
SKILL.mdUpdated Jun 3, 2026