agentsclimarketplace

Context checkpoint

Skill pipipip169/fable5-handoff/sources/adamentwistle-fable-skills/context-checkpoint

The retirement handoff of Claude Fable 5 - written by the model itself. Domain-neutral discipline files that transfer flagship working method to any model, any agent framework, any team.

Install
npx -y skills add pipipip169/fable5-handoff --skill context-checkpoint

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

  • 25 days oldThe repository was created 25 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Externalize working state to files so long tasks survive context compaction and interruption. Use at the start of any task likely to exceed ~30 minutes or span many files, and again whenever a major phase completes or the plan changes.

SKILL.md

1.7 KB, as published. Nobody here has run it

context-checkpoint

Context windows compact; whatever lives only in conversation history can silently vanish. Durable state lives in files.

  1. At task start, write a scratch plan file (in the session scratchpad or a .claude/-ignored path): the goal in one line, acceptance criteria, ordered steps, and decisions made. This is the file future-you re-reads after compaction instead of re-deriving everything.
  2. Update it at phase boundaries, not continuously: tick completed steps, record decisions with their why ("chose X over Y because Z" — the why is what evaporates first), note dead ends so they aren't retried.
  3. Record exact state for resumption: the failing command and its current output, the file:line you were editing, the next concrete action. "Continue the migration" is useless after compaction; "run pytest tests/test_ingest.py -k utf8, still failing on line 214 null check" resumes in seconds.
  4. After compaction or a fresh session, trust the file over the summary. Re-read the checkpoint and the current git diff before acting — the compacted summary is lossy; the file and the diff are ground truth.
  5. Commit at safe points when the work is committable — a described commit is the strongest checkpoint there is. Don't let 400 lines of uncommitted work ride on one fragile context.

Failure mode this prevents: the post-compaction session confidently redoing step 2 while overwriting completed step 4.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.