Coding

checkpoint

Try it

Part of the Overpowered skill suite. Create a durable, self-contained state package so a fresh agent, session, or different harness can resume non-trivial work without the original chat history. Use for long-running work, context resets, handoffs, or pauses; not for simple finished tasks.

What it does

Part of the Overpowered skill suite. Create a durable, self-contained state package so a fresh agent, session, or different harness can resume non-trivial work without the original chat history. Use for long-running work, context resets, handoffs, or pauses; not for simple finished tasks.

The skill document

Checkpoint

Core rule

Persist the state required to continue, not a transcript of how the conversation felt.

Procedure

Create or update a CHECKPOINT.md in the working directory unless the user specifies another location.

Capture:

  1. Goal — current objective and completion criteria.
  2. Current state — what is done, in progress, and not started.
  3. Authoritative inputs — exact artifacts/sources the receiver should trust.
  4. Decisions — decisions made and why.
  5. Rejected paths — only when retrying them would waste time.
  6. Artifacts — files, branches, queries, outputs, and their purpose.
  7. Evidence — checks already performed and their results.
  8. Known issues / unknowns — material unresolved items.
  9. Next action — the smallest useful next step.
  10. Environment assumptions — only non-obvious requirements.

Receiver test

Before finishing, ask: Could a capable fresh agent resume correctly from this file plus the referenced artifacts, without the chat transcript? If not, add what is missing.

Gotchas

  • Do not copy the full conversation.
  • Do not write vague state such as “mostly done.”
  • Do not omit negative decisions that a new agent is likely to repeat.
  • Do not claim verification that was not actually performed.

Stop condition

Stop when the receiver test passes.

Related skills

using-overpowered

Part of the Overpowered skill suite. Route knowledge-work and enterprise-automation tasks through the smallest relevant combination of Overpowered skills, and invoke gear-up only when a material execution capability is genuinely missing. Use when multiple suite skills may apply or when starting a non-trivial task involving organizational knowledge, conflicting evidence, evolving artifacts, business rules, process automation, risky side effects, verifiable completion, or an uncovered capability gap.

skillify

Part of the Overpowered skill suite. Generalize and package a qualified or otherwise proven repeatable workflow into a portable Agent Skill with concise instructions, progressive disclosure, trigger-focused metadata, gotchas, and evals. Use after real execution evidence exists—especially for a Skill Academy capability ready to graduate; do not use as the hot-generation mechanism during a task or merely to save a conversation.

completion-audit

Part of the Overpowered skill suite. Independently determine whether a claimed task outcome is actually complete by reconstructing completion criteria and checking fresh evidence. Use before declaring work done, accepted, migrated, reconciled, fixed, or delivered; not as a general planning loop.

automate-this

Part of the Overpowered skill suite. Transform a described human/business process into an executable automation design by separating deterministic steps, agentic reasoning, existing systems, and human gates; define triggers, data, decisions, exceptions, controls, and verification. Use when the goal is to automate a process, not merely understand it or interview the user about it.

gear-up

Part of the Overpowered skill suite. Create and activate the smallest temporary capability needed to close a real execution gap during a task, then measure whether it helped and discard or nominate it for reuse. Use only when installed skills, available tools, existing context, knowledge retrieval, and Academy candidates cannot reliably perform a material part of the objective; supports temporary skills, context instructions such as AGENTS.md fragments, and runtime tools when the harness can load them dynamically.

Check

Check (checkhq.com). Use this skill for ANY Check request — searching and reading data. Whenever a task involves Check, use this skill instead of calling the...

1 installs