overnight-mode
Overnight mode, the user away while admirals run on standing and night orders.
Skill by bodnarbalazs
Chalk Labs
Need help deploying overnight-mode?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/bodnarbalazs-overnight-mode
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/overnight-mode
What's inside
- Skill name
overnight-mode- Description
- Overnight mode, the user away while admirals run on standing and night orders. Load only on the chief of staff's muster or morning-brief announcement, or when the user calls overnight mode.
- License
- MIT
- Instructions
- 70 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Overnight Mode **Overnight mode** is any long absence of the user. At **muster**, the chief of staff verifies `overnight: on` with `dydo mode show` before announcing the start; at the **morning brief**, it verifies `overnight: off` before announcing the end. In between, **each admiral keeps the conn** of its Project and the chief of staff keeps only the **night watch**. Intent, a watch schedule and message delivery do not establish a mode transition. Admirals obey the **standing orders** below plus the **night orders**: one Project update per Project, set at muster. Night orders expire at the morning brief and win where the two conflict. ## Muster The already-invoked chief of staff uses its own resources below. Resolve each resource inside the installed `chief-of-staff` skill folder; read it directly without invoking another officer role. 1. While the user is present, complete the [approval preflight](resources/approval-preflight.md) for both startup and shutdown. Resolve the supported approval route or exact terminal fallback before proceeding; capability discovery is not a mode change or an approval-UI test. 2. Complete `chief-of-staff`'s `resources/muster.md`. Use `grilling` for decisions and `wizard` for terminal steps only. Proceed once every included admiral confirms its agreed work is runnable and every gap has the user's explicit disposition. 3. Use the preflight's route for `dydo mode set overnight on`. After actual execution, independently run `dydo mode show` and require `overnight: on`. On denial, failure or missing readback, report activation incomplete with the current observed state and exact human action; keep the user present and do not announce activation or say goodnight. 4. Create the 30-minute watch schedule, returning to this thread. Preserve the complete schedule ID exactly as returned by the host, including any prefix, for deletion at shutdown. If scheduling fails, report mode on with watch unavailable and resolve it while the user is present. 5. Announce the verified start to every included admiral. Done when readiness is confirmed, the schedule ID is preserved, mode is on and every included admiral has the announcement. ## Standing orders 1. **Keep the conn.** Project decisions are the admiral's. 2. **Decide by weight.** Easily reversible: decide and record `DECIDED`. Other calls: `what-would-balazs-do`, whose trigger map decides it or prepares the Question; irreversible calls are always a Question. 3. **Merge into the feature branch** through your captain's Merge chain, which runs `dydo merge`; never into master or main. 4. **The board is the report.** Send the chief of staff no status. Contact it for cross-Project trouble: collisions, shared-resource problems, and heads-ups or permission before actions that affect other agents. It relays down admiral → captain → crew. Answer a nudge in one line. 5. **Keep moving.** One blocker never stops the night: take the next pickable work. Stop only at the night orders' stop line or when nothing is pickable. The walking skeleton does not wait either: its Walkthrough Ticket blocks nothing; post `DECIDED` on it, carry on, and treat a wrong direction as a revert. 6. **Write for a reader who wasn't there**, every Question and `DECIDED`. Withdraw only Questions made obsolete by discovered facts or completed work; keep tribunal-decided Questions awaiting ratification. ## Night watch The chief of staff runs its Overnight duties watch step every 30 minutes, using its `resources/night-survey.md`. The surveyor inspects; the chief of staff acts on its findings: 1. Idle, or waiting with nothing in flight and runnable work: send one specific nudge naming the evidence. 2. Nudge unanswered: inspect the relevant captain threads and intervene with the owner. The board and threads carry state; never ask for routine STATE. Mediate cross-Project trouble by chief-of-staff Method step 4, Mediate a collision, relaying heads-ups down the chain. After any gap in the watch, rebuild from Linear and sweep everyone by Method step 5, Sweep the board. ## Morning brief When the user says he's back, in the chief of staff's conversation: 1. Complete the [approval preflight](resources/approval-preflight.md) for shutdown while the user is present; recheck if host or permission mode changed overnight. 2. Delete the watch schedule using the complete preserved ID. List schedules and confirm that exact ID is absent before claiming the watch stopped; report a deletion or verification gap. 3. Settle any pending survey through its host lifecycle so no tick remains in flight. 4. Use the preflight's route for `dydo mode set overnight off`. After actual execution, independently run `dydo mode show` and require `overnight: off`, then announce the end to every included admiral. On denial, failure or missing readback, report shutdown incomplete with the current observed state and exact human action. A deleted schedule does not prove mode is off. 5. Fill `chief-of-staff`'s `resources/morning-brief.md` in the conversation, writing nothing to the repo. Use `wizard` for terminal unblockers only. Done when the schedule is confirmed absent, pending surveys are settled, mode is off, admirals have the announcement and the user has the brief.
From bodnarbalazs/dydo, MIT.
More from bodnarbalazs/dydo
- 0.622score
maintain-formal-verification-skill
bodnarbalazs
Bring a project's formal verification back in line with the code after the code changes.
SkillAFreshness: A★ 6Claude Code +9 - 0.622score
maintain-verification-skill
bodnarbalazs
Keep a project's verification skill and feature map true to the running app.
SkillAFreshness: A★ 6Claude Code +9 - 0.622score
co-thinker
bodnarbalazs
An idea not ripe yet.
SkillAFreshness: A★ 6Claude Code +9 - 0.622score
retro
bodnarbalazs
Kaizen for the harness.
SkillAFreshness: A★ 6Claude Code +9 - 0.619score
wizard
bodnarbalazs
Human-only setup, credentials or cutover steps need a repeatable interactive bash procedure; the Enablement captain uses this method.
SkillAFreshness: A★ 6Claude Code +9 - 0.617score
in-your-own-words
bodnarbalazs
Restate in your own words what you think the user's goals are and what problem they're trying to solve.
SkillAFreshness: A★ 6Claude Code +9 - 0.608score
bro
bodnarbalazs
Stop.
SkillAFreshness: A★ 6Claude Code +9 - 0.597score
ketchup
bodnarbalazs
Catch the user up on what changed in this chat since their last message, in a short reply built on real visuals.
SkillAFreshness: A★ 6Claude Code +9 - 0.597score
domain-modeling
bodnarbalazs
Build and sharpen a project's domain model.
SkillAFreshness: A★ 6Claude Code +9 - 0.597score
codebase-design
bodnarbalazs
Shared vocabulary for designing deep modules.
SkillAFreshness: A★ 6Claude Code +9 - 0.597score
diagnosing-bugs
bodnarbalazs
Diagnosis loop for hard bugs and performance regressions.
SkillAFreshness: A★ 6Claude Code +9 - 0.597score
prototype
bodnarbalazs
Build a throwaway prototype to answer a design question.
SkillAFreshness: A★ 6Claude Code +9
Signals
- Stars
- 6
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 22, 2026 (13 days ago)
- Last release
- Oct 19, 2026 (17 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/bodnarbalazs/dydo
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from bodnarbalazs
- 0.622score
maintain-formal-verification-skill
bodnarbalazs
Bring a project's formal verification back in line with the code after the code changes.
SkillAFreshness: A★ 6Claude Code +9 - 0.622score
maintain-verification-skill
bodnarbalazs
Keep a project's verification skill and feature map true to the running app.
SkillAFreshness: A★ 6Claude Code +9 - 0.622score
retro
bodnarbalazs
Kaizen for the harness.
SkillAFreshness: A★ 6Claude Code +9 - 0.622score
co-thinker
bodnarbalazs
An idea not ripe yet.
SkillAFreshness: A★ 6Claude Code +9 - 0.619score
wizard
bodnarbalazs
Human-only setup, credentials or cutover steps need a repeatable interactive bash procedure; the Enablement captain uses this method.
SkillAFreshness: A★ 6Claude Code +9 - 0.617score
in-your-own-words
bodnarbalazs
Restate in your own words what you think the user's goals are and what problem they're trying to solve.
SkillAFreshness: A★ 6Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get overnight-mode running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
overnight-mode skill. Agent Mart. https://agentmart.dev/skills/bodnarbalazs-overnight-mode (last updated 2026-10-24)