agentleFS
arbelharbelh/perdure/skills/continue/SKILL.md

continue

Skill0 starsChanged 28 days agoMITView on GitHub

Resume a previously started perdure workflow. Reads the workflow's

25 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use dependency-watch, perform-task and 7 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • telegramdesktopSkill

    dependency-watch

    Audit Telegram Desktop dependencies on freshly fetched origin/dev for releases and security fixes, including upstream lag and backport candidates in patched forks. Use for the scheduled dependency monitor or an explicitly requested dependency report.

    Also in 5 projects that use this

  • telegramdesktopSkill

    perform-task

    Resolve, start or resume, implement, review, test, and publish exactly one existing ai-tdesktop task by short slug or full dated id, including rare blocked retries and split-required results. Use when the user invokes $perform-task or /perform-task with a known task name, or when the continue scheduler delegates one selected task. Runs standard review lenses with fast applicability bailouts and selects task-specific domain and evidence instruments without selecting additional work.

    Also in 5 projects that use this

  • JerovinoSkill

    process-inbox

    Processes tasks.md inbox items by triaging each one to its proper location in the vault

    Also in 5 projects that use this

  • raineSkill

    rebase

    Rebase the current branch with smart conflict resolution.

    Also in 6 projects that use this

  • lepinkainenSkill

    code-health

    Analyze codebase health - large files, test coverage gaps, duplicate code, dead/legacy code, and documentation issues. Use when asked to scan, audit, or assess code quality, find refactoring targets, or identify technical debt.

    Also in 3 projects that use this

  • BlockSkill

    plan

    Draft a TDD, PRD, or lightweight plan in the conversation without committing it.

    Also in 4 projects that use this

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. 0 are common, 3 are distinctive, and 11 are very specific to this project.

  • 0

    Common

    Most of the 23 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 11

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If no argument, scan perdure/workflows/*.md for active workflows and ask which to resume.

    line 20
  2. Very specific:

    If perdure/HANDOFF.md holds merge-conflict markers (a line starting with <<<<<<< , then ======= or >>>>>>> ), read each side on its own: …

    line 25
  3. Distinctive:

    When a side lacks the stamp, someone wrote it by hand.

    line 34
  4. Very specific:

    If $1 provided: find perdure/workflows/*<$1>*.md that matches (also check <YYYY>/ subdirectories for aged terminal files)

    line 38
  5. Distinctive:

    If multiple match: show options and ask user to pick

    line 39
  6. Distinctive:

    If none match: list the closest-named files and suggest alternatives

    line 40
  7. Very specific:

    Read the full workflow file (the schema is docs/RECORDS-CONVENTION.md, also at references/RECORDS-CONVENTION.md beside this SKILL.md in a …

    line 43
  8. Very specific:

    If status: completed or status: abandoned: warn the user and ask whether to reopen (changes status back to in-progress) or just display the …

    line 48
  9. Very specific:

    If status: in-progress: proceed with resumption

    line 50
  10. Very specific:

    If # Plan still holds the scaffold placeholder or has no items, write the plan now from the goal, the diff and what is left, before …

    line 53
  11. Very specific:

    Update the updated: timestamp in the workflow file's frontmatter

    line 101
  12. Very specific:

    If status was changed from completed/abandoned → in-progress, note the reopen under # Decisions AND annotate the existing # Review verdict …

    line 102
  13. Very specific:

    After resumption, continue from # Next step however this session works — inline, or through whatever agents the user has.

    line 112
  14. Very specific:

    Graceful missing file: if no workflow matches, say so clearly and suggest /perdure:list to list available ones

    line 125

“Common” is measured: kept by at least 6 in 10 of the 23 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Resume a Workflow

You are restoring a workflow's state so work can continue.

Arguments

  • $1 — the workflow slug (or partial match). Example: add-jwt-refresh
  • If no argument, scan perdure/workflows/*.md for active workflows and ask which to resume.

Steps

0. Check the HANDOFF page

  • If perdure/HANDOFF.md holds merge-conflict markers (a line starting with <<<<<<< , then ======= or >>>>>>> ), read each side on its own: the lines outside the conflicts plus that side's half of each one.
  • Regenerate the page only when every side carries the generator's stamp, as the close-gate hook does: the side starts with # HANDOFF, and one of its next four lines starts with > Derived orientation page, generated. Say so in one line: the page is derived, so a conflict is never resolved by hand. python3 "${CLAUDE_PLUGIN_ROOT}/scripts/handoff.py" . (step 3b says where the scripts are without ${CLAUDE_PLUGIN_ROOT})
  • When a side lacks the stamp, someone wrote it by hand. Leave the page as it is, say so in one line, and go on with the resume. Do not ask.

1. Locate the workflow file

  • If $1 provided: find perdure/workflows/*<$1>*.md that matches (also check <YYYY>/ subdirectories for aged terminal files)
  • If multiple match: show options and ask user to pick
  • If none match: list the closest-named files and suggest alternatives

2. Read the workflow file

  • Read the full workflow file (the schema is docs/RECORDS-CONVENTION.md, also at references/RECORDS-CONVENTION.md beside this SKILL.md in a skills-CLI install; real files may be freeform — read what's there)

3. Check status

  • If status: completed or status: abandoned: warn the user and ask whether to reopen (changes status back to in-progress) or just display the state
  • If status: in-progress: proceed with resumption

3b. Check the plan

  • If # Plan still holds the scaffold placeholder or has no items, write the plan now from the goal, the diff and what is left, before continuing; say so in the resume message. The lint reports this as placeholder.
  • To see it, run the lint on the record: python3 "${CLAUDE_PLUGIN_ROOT}/scripts/rot-lint.py" <workflow-file> (${CLAUDE_PLUGIN_ROOT} is the plugin directory, set by Claude Code. Without it, the same script sits in the plugin checkout's root scripts/ directory, or in scripts/ beside this SKILL.md in a skills-CLI install.)

4. Re-inject state into conversation

Present to the user and load into context. The first line is always the resume line: the task in a few words (the slug, or the goal shortened) and the record's # Next step:

Resumed from perdure: <task>, next: <next step>

## Resuming workflow: <slug>

**Goal**: <goal>
**Started**: <started> · **Last updated**: <updated>

### Plan progress
<plan section, with checkboxes showing completion>

### Key findings so far
<findings section summary>

### Key decisions made
<decisions section summary>
(Each holds as written. A later record that changed one names this record on
its line; if another record disagrees and neither names the other, hold both
as unresolved rather than picking by tone.)

### Files modified so far
<files modified section>

### Open questions
<open questions section>

### Next step
<next step from the file>

---
Ready to continue? I'll pick up from "<next step>".

5. Update metadata

  • Update the updated: timestamp in the workflow file's frontmatter
  • If status was changed from completed/abandoned → in-progress, note the reopen under # Decisions AND annotate the existing # Review verdict block as (pre-reopen) — re-completing this workflow requires a NEW verdict (or a new recorded skip) covering the post-reopen work, per docs/RECORDS-CONVENTION.md. (Terminal files are normally immutable — prefer a new file, adding a continued-by: link to the old one per the convention; reopen only when the user explicitly chooses it. The convention's other exception is a decision renumbering, in the decide skill's step 5.)

6. Continue the work

After resumption, continue from # Next step however this session works — inline, or through whatever agents the user has. Do not start editing on a placeholder plan: step 3b writes one first. Close through the gate when done (/perdure:verdict, then /perdure:close).

Safety rules

  • Never modify content of Findings, Decisions, Files modified sections — those are historical record. Two exceptions: the close pass's append-only promotion annotations, and a decision renumbering after two branches took one number (the decide skill's step 5)
  • Always show the user what's being restored before injecting into context
  • Respect completed/abandoned status — reopening requires explicit user confirmation
  • Graceful missing file: if no workflow matches, say so clearly and suggest /perdure:list to list available ones

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.