Clear session and start fresh
View or modify heartbeat settings
Show Discord bot status and manage global session
Show heartbeat plugin help
Create, list, edit, or delete cron jobs. Triggers: create a job, add a job, new job, schedule a task, schedule a prompt, set up a cron, automate, run on a schedule, recurring task, periodic task, timed task, I want to schedule, I want to create a job, add scheduled task, manage jobs, job list, delete job, remove job, edit job, run job
Create new skills for Claude Code. Use when users ask to create a skill, add a skill, make a new command, build a skill, add a slash command, create a plugin skill, or define a new automation. Trigger phrases include "create a skill", "new skill", "add a skill", "make a command", "build a skill", "I want a skill that", "add slash command", "create automation".
Explicitly escalate a Datadog On-Call page (or proactively page on-call) when an automated first-responder cannot safely resolve an alert. Bumps an existing page to the next escalation tier, or creates a new page targeting the on-call team, so a human is engaged now instead of waiting for the next timed tier. Use when an agent gives up, is unsure, or the fix did not recover the monitor. Trigger phrases include "escalate the page", "escalate to on-call", "page a human", "notify on-call", "hand off to on-call", "I can't fix this alert", "escalate the incident".
Re-check whether a firing Datadog monitor / the metric that tripped it has recovered, after attempting a remediation. Use when an incident-response agent needs to confirm a fix worked before resolving a page, or to decide whether to escalate. Trigger phrases include "did the alert recover", "re-check the monitor", "is the metric back to normal", "confirm recovery", "check monitor state", "has it recovered", "verify the fix cleared the alert".
Explicitly dismiss/resolve a Datadog On-Call page after confirming the underlying monitor has recovered. Use when an incident-response agent has fixed the problem and needs to close out the page rather than leave it for a human or rely on auto-resolve. Trigger phrases include "resolve the page", "dismiss the alert", "close the page", "clear the alert", "mark resolved", "the fix worked resolve it", "dismiss the datadog page".
Search and install skills from skills.sh and GitHub repos. Use when users ask to find skills, install skills, download skills, add skills from GitHub, search for skills, browse skills, get a skill, or want new capabilities. Trigger phrases include "install skill", "find skill", "search skills", "add skill", "download skill", "get skill from github", "skills.sh", "browse skills", "what skills are available", "I need a skill for".
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimnpx claudepluginhub northisup/errandd --plugin erranddBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A self-hosted daemon that puts your coding agent on watch — listening to your dev and prod environment and acting on what it sees.
Errandd runs a coding-agent CLI — Claude Code by default, or Pi — as a background daemon wired into the systems your team already uses. It listens for events (GitHub pull requests, Sentry errors, Datadog alerts, Linear issues) and runs scheduled routines (morning ticket triage, nightly sweeps, recurring checks). Each event or tick is matched against your routines — plain markdown files, version-controlled — and handed to the agent to investigate, triage, or fix. A web dashboard and Telegram/Discord/Slack bridges let you watch and steer it in real time.
It's built on the coding-agent subscription you already pay for — no per-call API bill, no cloud to trust with your source. (Formerly clawdcode.)
GitHub · Sentry · Datadog · Linear ─┐
├─► match routines ─► enqueue job ─► agent (Claude / Pi) ─► investigate · triage · act
cron schedule ──────────────┘
on: triggers.Errandd exposes one signed webhook per provider. Each incoming delivery is verified, recorded, and matched against your routines' triggers.
| Source | Endpoint | Auth | Triggers |
|---|---|---|---|
| GitHub | POST /api/webhooks/github | HMAC X-Hub-Signature-256 — ERRANDD_GITHUB_WEBHOOK_SECRET | prs, pr, comments, reviews, checks, issues |
| Sentry | POST /api/webhooks/sentry | HMAC Sentry-Hook-Signature — ERRANDD_SENTRY_CLIENT_SECRET | sentry |
| Datadog | POST /api/webhooks/datadog | X-Errandd-Token / ?token= — ERRANDD_DATADOG_WEBHOOK_SECRET | datadog |
| Linear | POST /api/webhooks/linear | HMAC Linear-Signature — ERRANDD_LINEAR_WEBHOOK_SECRET | linear |
The GitHub receiver understands the real payloads — it parses pull requests, review comments, checks, and issues, skips its own bot activity, honors a claw:ignore label, and drops CI noise before it ever reaches the agent. Sentry keeps a first-seen ledger so you're triaged once per issue, not once per event. Datadog ships a recommended payload template you paste into your monitor.
Prefer to push events yourself? POST /api/inject runs a one-shot agent message from any script or CI step (bearer/web token, or a static settings.apiToken).
A routine is a markdown file describing a task, with an on: block declaring what triggers it. Point Errandd at a git repo (jobsRepo.url / ERRANDD_JOBSREPO_URL) and every top-level .md file becomes a routine — cloned on start, pulled on an interval. Your automation lives in version control, reviewed like any other code.
---
description: Review new and updated pull requests
on:
- prs: true
---
Review the pull request diff for bugs, missing tests, and unclear code.
Leave inline review comments where you can be specific, then post a short
summary. Approve only if it's clearly safe.
Triggers combine freely — an event, a schedule, or both:
---
description: Triage new Sentry issues, and sweep the backlog each morning
on:
- sentry: true
- schedule: "0 9 * * 1-5"
---
For each Sentry issue, find the offending code, assess severity and likely
cause, and open a Linear ticket with a proposed fix when it's actionable.
NorthIsUp's Utils — statusline + helpers for Claude Code: org/repo + branch + PR/CI/review state on the left, context bar / quota / model on the right.
Adam's personal skills, agents, and commands
Save & auto-restore tmux sessions with Claude Code --resume per pane. Restores windows, panes, cwd, and re-launches each pane's previous claude session by id.
Ultra-compressed communication mode. Cuts 65% of output tokens (measured) while keeping full technical accuracy by speaking like a caveman.
Comprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Standalone image generation plugin using Nano Banana MCP server. Generates and edits images, icons, diagrams, patterns, and visual assets via Gemini image models. No Gemini CLI dependency required.
Multi-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
Unified capability management center for Skills, Agents, and Commands.