Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/LFTPadilla/agent-dev-kitnpx agentmods add skills/lftpadilla/agent-dev-kit/overnight-taskWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/lftpadilla/agent-dev-kit/overnight-task)<a href="https://agentmods.dev/skills/lftpadilla/agent-dev-kit/overnight-task"><img src="https://agentmods.dev/badge/skills/lftpadilla/agent-dev-kit/overnight-task/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/lftpadilla/agent-dev-kit/overnight-task"><img src="https://agentmods.dev/badge/skills/lftpadilla/agent-dev-kit/overnight-task.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00065 | $0.00761 |
| Opus 5.5 | $0.00026 | $0.00304 |
| Sonnet 5.5 | $0.00013 | $0.00152 |
| Haiku 4.5 | $0.00006 | $0.00076 |
Grade A, and why
overnight-task scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.
overnight-task
Use this skill when the user explicitly says they are leaving a task running for hours, overnight, or unattended and asks the agent not to stop for clarifying questions.
Invocation Signals
- "tienes toda la noche"
- "no me hagas más preguntas"
- "no voy a estar disponible"
- "deja esto funcionando"
- "apenas termines puedes apagar..."
- "necesito un reporte completo para mañana"
- "audita todo lo que encuentres"
Do not invoke for short, interactive, or low-risk tasks where the user is present.
Operating Contract
- No questions after activation. Make conservative decisions and document them.
- Plan before execution.
- Keep a journal of material actions and decisions.
- Checkpoint every 2-3 tasks or before risk increases.
- Verify each completed task.
- Do not commit, push, deploy, mutate production, touch secrets, or shut down machines unless the user explicitly authorizes that action in the same session.
- Produce a final report with what changed, what was verified, what remains open, and what the next session should review.
Recommended Runner
Initialize the run directory:
node overnight-task-kit/scripts/overnight-runner.mjs init --title "<short-title>"
Then use the generated files:
SPEC.md: falsifiable scope and ambiguity assessment.PLAN.md: vertical slices with acceptance criteria and verification commands.JOURNAL.md: chronological record.CHECKPOINTS.md: stop/continue gates.REPORT.md: final handoff.
Mandatory Flow
- Research: inspect local docs, code, backlog, and relevant external docs if the task depends on current APIs or packages.
- SPEC: write requirements, non-goals, risks, assumptions, and ambiguity score.
- PLAN: split into small independently verifiable tasks.
- Execute: complete one task at a time; update plan status as work lands.
- Verify: run tests, linters, smoke checks, or manual checks appropriate to the task.
- Report: fill
REPORT.mdand include plan deviations. - Shutdown handoff: only if the user authorized a shutdown in this session.
Follow
references/shutdown-sequence.md: run the exact command they gave, never fire while the run still has work of its own, and decide "finished" from agent activity, not from process existence. Record the evidence from that reference in the final report.
What ships with it
10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- overnight-task.opencode.md 6.4 KB
- references/agent-quickref.md 4.1 KB
- references/connections.md 510 B
- references/learnings-template.md 694 B
- references/plan-template.md 287 B
- references/planning-and-task-breakdown-canonical.md 6.8 KB
- references/planning-protocol.md 1019 B
- references/session-learnings-template.md 239 B
- references/shutdown-sequence.md 2.9 KB
- references/spec-template.md 473 B
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago Changed · +3 lines f6ff4bfeaa3a
- 1mo ago First seen · 93 lines · 65 tokens per session scan A 5b977d88c50f
overnight-task is a skill published in the GitHub repository LFTPadilla/agent-dev-kit (2 stars, last pushed 21d ago), licensed MIT. It adds 65 tokens to every session and 761 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
simplify-code
Parallel 4-agent cleanup of recent code changes.
community-research-insight
Extract structured insight briefs from community research transcripts or notes. Produces pain points, stakeholder needs, opportunity maps, risks, and follow-up questions. Requires human review before publication.
output-eval-error-analysis
Systematically review workflow traces to identify failure modes before building evaluators. Use when starting an eval project, after significant pipeline changes, or when production quality drops.
install-jazz-github-action
Set up Jazz as a GitHub Action for automated PR code review and on-demand /jazz PR assistance. Use when the user asks to add Jazz to CI, create a GitHub Action workflow, set up PR review automation, enable /jazz commands on PRs, or integrate Jazz with GitHub.
chief-of-staff
The CEO's right hand — maintains validated project context, prepares task contracts, logs cross-department decisions, and runs the weekly review. Use when the user says "what's in flight", "resume the project", "log this decision", "weekly review", "turn this idea into a brief", or when multi-department work needs…
appsec-review
Produces a bounded defensive application security review: threat model, security-sensitive diff, dependencies, and secrets hygiene with source-grounded findings. Use when the user says "security-review this release", "check this integration before shipping", "review the auth change", or "audit our agent's data access".