Workspace resume
Skill Dokuganrryu/ai-agent-skills-and-subagents/skills/workspace-resume
Use when resuming work from an old agent session, projectless workspace, handoff folder, dirty repo, pointer AGENTS.md, CURRENT_WORK ledger, or after context loss/compression, especially when the user wants the real repo/current state/next slice recovered before editing.From its SKILL.md
npx -y skills add Dokuganrryu/ai-agent-skills-and-subagents --skill workspace-resumeAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 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.
- 0 stars0 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.
SKILL.md
2.7 KB, 533 tokens by cl100k_base, as published. Nobody here has run it
Workspace Resume
Use this skill when the main risk is starting from the wrong folder, stale context, or an old plan instead of the actual current workspace.
Success Criteria
- The real workspace or repo is identified.
- Pointer folders, handoff files, and current ledgers are separated from source-of-truth files.
- Dirty worktree state and likely user changes are preserved.
- The next actionable slice is recovered without re-planning the whole project.
- Verification run or remaining verification gap is named.
Workflow
-
Pin the requested context
- Identify the folder, repo, thread, task name, or date the user referenced.
- Do not assume the current folder is the real repo.
-
Read pointer and state files first
- Check
AGENTS.md,README*,CURRENT_WORK*,HANDOFF*,KEY_DECISIONS*, task ledgers, package indexes, and transfer artifacts. - Note which files are coordination notes and which are source-of-truth project files.
- Check
-
Locate the real workspace
- Inspect the repo root, branch, recent commits, and
git status. - If the folder is a pointer or export package, follow the pointer before editing.
- Never revert unrelated changes.
- Inspect the repo root, branch, recent commits, and
-
Recover the current task state
- Use recent session names, memory, rollout summaries, and local ledgers only as much as needed.
- Prefer exact local evidence over broad memory.
- Extract the latest accepted decision, current blocker, and next smallest slice.
-
Verify cheaply when useful
- Run existing focused tests, build checks, linters, or smoke checks when they are safe and relevant.
- If verification is too expensive or blocked, say exactly what remains unverified.
-
Report before changing code
- Name the real repo, authoritative files, dirty-state warning, recovered next slice, and verification status.
- Then proceed with edits only if the user's request calls for action.
Output Template
**Recovered State**
- Real workspace:
- Source-of-truth files:
- Current branch/state:
- Latest accepted decision:
- Next slice:
- Verified:
- Unverified:
Common Traps
- Treating a projectless agent folder as the source repo.
- Restarting discovery from scratch when a ledger already names the next slice.
- Using memory as current fact without checking cheap local evidence.
- Reverting or cleaning up user changes while trying to get a clean state.
- Producing a long handoff when the user needs the next action.
What ships with it: 1 file
66 B alongside SKILL.md
agents/
- openai.yaml66 B
Gives 0 of the 12 instructions most hr recruiting skills give in 533 tokens
Counted across 356 of the 357 authors here whose files we hold, read 2026-08-07
- Quantify achievements with specific metricsin 14 of 356, across 6 files
- Keep the resume under two pagesin 14 of 356, across 6 files
- Request the full job description if not providedin 12 of 356, across 4 files
- Extract keywords and prioritize job requirementsin 12 of 356, across 4 files
- Stop and ask for clarification if required inputs are missingin 12 of 356, across 5 files
- Map candidate experience to job requirementsin 11 of 356, across 3 files
- Ask if the user wants adjustmentsin 11 of 356, across 3 files
- Provide strengths and gap analysis after the resumein 10 of 356, across 2 files
- Request candidate background details if not providedin 10 of 356, across 2 files
- Format experience bullets as action verb plus resultin 10 of 356, across 2 files
- Ask for missing inputs before startingin 10 of 356, across 9 files
- Use exact job description terminologyin 9 of 356, across 1 file
Said here and by no other author read
- identify the referenced folder repo or task
- do not assume the current folder is the real repo
- separate coordination notes from source-of-truth files
- inspect repo root branch and git status
- follow pointers before editing
- prefer exact local evidence over broad memory
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.