Drewgent configuration for opencode — agent profiles, skills, scripts, and tooling
This repository also configures its own agents. See what opencode-drewgent tells them →
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Skill "kanban-stuck-task-recovery" from humanerd-drew/opencode-drewgent, covering kanban stuck task diagnosis & recovery, when to use, diagnosis, 1. find all inprogress tasks with their worker status and 2. check if worker pids are alive.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Skill "launchd-process-health-check" from humanerd-drew/opencode-drewgent, covering launchd process health check, problem, always verify with these 3 commands, 1. launchd's view (may be stale) and 2. is it actually listening? (ground truth).
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Clean up {{AGENTNAME}}'s persistent memory file (MEMORY.md) when it nears the 8K char cap. Identify resolved/one-time entries, preserve operational facts, verify after.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex needs its repo
Skill "ontology-query" from humanerd-drew/opencode-drewgent, covering ontology-query — opencrab graph query tool, script location, commands, 1. list all nodes and 2. show space distribution.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
A convention for managing persistent knowledge when an OpenCode knowledge plugin provides tools such as remember(), recall(), and memory-stats(). It defines when to retrieve context, save decisions, and avoid storing secrets.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Filename uniqueness, wikilink resolution safety, P-layer identity in naming, and Obsidian exclusion boundaries. Prevents duplicate-name ambiguity and graph pollution.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex needs its repo
Patterns, pitfalls, and workflows for local Cloudflare Workers development with D1, static assets, multi-source integration, and the m-log refactoring architecture.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
A monitoring setup that watches existing work or outside sources and turns selected topics into batches of content, such as a blog draft, an X thread, and a LinkedIn post.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
A content-production agent that watches work activity and turns selected work into blog posts, visual files, and social-media drafts, with WordPress publishing support.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Dark-themed SVG architecture/cloud/infra diagrams as HTML.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
A generator for infographics, or visual summaries of information, with 21 layout types and 21 visual styles.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Design one-off HTML artifacts (landing, deck, prototype).
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex needs its repo
Generate images, video, and audio with ComfyUI — install, launch, manage nodes/models, run workflows with parameter injection. Uses the official comfy-cli for lifecycle and direct REST/WebSocket API for execution.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Author/validate/export Google's DESIGN.md token spec files.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Humanize text: strip AI-isms and add real voice.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
A Korean-language editing skill that spots common signs of AI-written text and rewrites the wording and rhythm to sound more natural. It keeps the original meaning, facts, numbers, names, and quotations unchanged.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Use when building creative browser demos with @chenglou/pretext — DOM-free text layout for ASCII art, typographic flow around obstacles, text-as-geometry games, kinetic typography, and text-powered generative art. Produces single-file HTML demos by default.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Throwaway HTML mockups: 2-3 design variants to compare.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Diagnose and recover cron jobs in jobs.json that stopped running (nextrunat=null, no recent execution). Distinguishes scheduler bug vs script quality issue. Applies 60-second recovery patch.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
4-tab insight-driven dashboard. Load this skill when the user asks about the agent dashboard — adding features, fixing display issues, debugging pusher collectors, or expanding tab content.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Add a script field to jobs.json + scheduler.py branch to bypass LLM for simple shell cron jobs (cost + reliability). Reusable pattern for any 'Run: python3 xxx.py' style cron.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Decomposition playbook + anti-temptation rules for an orchestrator profile routing work through Kanban. The "don't do the work yourself" rule and the basic lifecycle are auto-injected into every kanban worker's system prompt; this skill is the deeper playbook when you're specifically playing the orchestrator role.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Pitfalls, examples, and edge cases for Hermes Kanban workers. The lifecycle itself is auto-injected into every worker's system prompt as KANBANGUIDANCE (from agent/promptbuilder.py); this skill is what you load when you want deeper detail on specific scenarios.
humanerd-drew/opencode-drewgent
Skill Claude CodeCodex
Skill "minimax-usage" from humanerd-drew/opencode-drewgent, covering minimax token plan — terminal usage check, 1. api endpoint (discovered 2026-06-02), 3. ready-made script (created 2026-06-02), 4. shell alias (user must add manually) and 5. pitfalls.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: