agentleFS
gregmosgregmos/memoforge/skills/continue/SKILL.md

continue

Skill38 starsChanged 48 days agoMITView on GitHub

Resume a memoforge legal-memo task and answer its open gate. Use only when explicitly invoked via /memoforge:continue, optionally with a task_id and a gate reply (proceed, continue, approve, edit: …, cancel, or answers like 1A 2C).

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 9 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

  • AnthropicSkill

    start

    Set up your bio-research environment and explore available tools. Use when first getting oriented with the plugin, checking which literature, drug-discovery, or visualization MCP servers are connected, or surveying available analysis skills before starting a new project.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 23 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

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

    If the reply contains the word cancel (ТЗ §2.4 d), run mf task cancel --workdir W immediately, print one line confirming the cancellation, …

    line 32
  2. Very specific:

    If the answer carries notice (D-242: the reply named the removed Brief mode), print notice verbatim first, then carry on as the rest of the …

    line 49
  3. Distinctive:

    If the reply is empty, or the task is not at a gate, skip this step — the loop below re-issues whatever is open.

    line 53

“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

memoforge / continue — re-entry into the router

Same protocol as /memoforge:memo: read skills/memo/references/router.md now and follow it, plus the authority and untrusted-content rules of skills/memo/SKILL.md. This file only covers getting back into the loop.

Every reply to the user is in the task's ui_language (D-178): text/questions/reprompt arrive localized from the CLI — print them verbatim; translate the English chat_line, the terminal response and error explanations yourself. Never translate tokens in backticks, file paths, commands or the text_fallback reply forms.

1. Split $ARGUMENTS

The first token is a task_id only if it looks like one (memo-…). Everything else is the reply — it may be empty.

2. Resolve the task

mf task resolve [task_id] --workdir <dir>   # both optional; no id = last unfinished task
  • {"unsupported": true, …} → print hint ("task from v1 — finish it with memoforge 1.1.1 or start a new task") and END.
  • {"errors": ["no_unfinished_task"]} → print No unfinished memoforge task. Start one with /memoforge:memo "<question>". and END; task_not_found → print it and suggest /memoforge:status.
  • Otherwise keep work_dir as W and note current_phase, gate, cancel_requested.
  • The descriptor carries no language: read it once with mf state get --workdir W --path ui_language and answer in it from here on — path_not_found: ui_language is a task from before the option, which means English; it is a business answer, never a CLI failure, so do not retry or finalize on it.

3. cancel first — before looking at the phase

If the reply contains the word cancel (ТЗ §2.4 d), run mf task cancel --workdir W immediately, print one line confirming the cancellation, then continue at step 5: next issues no new work and walks the task to its finalized terminal step.

4. A reply to an open gate

  • At the plan gate an Edit answer that asks for another memo language is not an edit for the planner: run mf task language --workdir W --memo <code>, then re-issue the gate with mf next. The error language_locked means the plan was already approved — tell the user the memo language is fixed.
  • This case never goes to gate parse — check it first, before the parser step below.

If the reply is non-empty, is not that memo-language case, and the task is at a gate ("gate": true), call mf next --workdir W to learn the open gate step — do not print its prompt again — and take step_id, attempt, generation from it. Hand the reply to the parser, which also closes the step:

mf gate parse --workdir W --step s-002 --attempt 1 --generation 0 --text "<reply verbatim>"

This is the only route for a text reply, kind: gate-text and kind: gate-auq alike (ТЗ D-34): for an AUQ gate at generation 0 the command performs the channel switch itself, writes gate_channel_switched and then parses the reply. Never convert the reply into AUQ answers by hand and never switch the channel yourself.

--generation is mandatory — pass the value next just returned; an answer at another generation is rejected as stale_generation.

If the answer carries notice (D-242: the reply named the removed Brief mode), print notice verbatim first, then carry on as the rest of the answer says.

A {"errors": […], "reprompt": "<text>"} answer is a business outcome, not a CLI failure (D-72): the reply was not understood or only partly understood and nothing is recorded. Print reprompt verbatim and END the turn — never repeat the command and never finalize; the user's next reply re-enters here.

If the reply is empty, or the task is not at a gate, skip this step — the loop below re-issues whatever is open.

5. Run the loop

Continue exactly as router.md describes: mf next --workdir W → render chat_line in the task's ui_language → act on kind → mf report where the kind requires it → next again, until a gate-text or terminal step ends the turn.

An answer that carries a dashboard key is handled by router.md §2a — one Artifact call per step (publish once, then write_db) — with file_path from the block and if_version = the version the previous write reported; version_mismatch → one get, one resend. It is Step 0 of that answer: make the call before acting on kind (before the Agent dispatch, before command[], before AskUserQuestion, before a terminal message), never after, and never inside the same Bash call as next — it is an Artifact tool call, not a Bash command. An error there is never a step failure.

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.