Skip to content
Skills Library
AI & agents · Automation

open-edit-close-execute

/open-edit-close-execute

Open an existing TASK, PROMPT, INTENT, or other input for editing, wait for close and supplied checks, then execute its saved requirements within caller authorization. Use when asked for `open task`, `open TASK`, `open TASK.md`, `edit task`, `edit TASK`, and `edit TASK.md`, or the same workflow for another input file.

Licence
MIT
Last repository change
11.11.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
6
Scripts
1
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 11.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What to check before installing
  • 1 scripts in this package. Review them before installing, as Claude may execute them.

Scripts: scripts/investigation_budget.py

What it does and when to use it

Open an existing TASK, PROMPT, INTENT, or other input for editing, wait for close and supplied checks, then execute its saved requirements within caller authorization. Use when asked for `open task`, `open TASK`, `open TASK.md`, `edit task`, `edit TASK`, and `edit TASK.md`, or the same workflow for another input file.

Original author description

Open an existing TASK, PROMPT, INTENT, or other input for editing, wait for close and supplied checks, then execute its saved requirements within caller authorization. Use when asked for `open task`, `open TASK`, `open TASK.md`, `edit task`, `edit TASK`, and `edit TASK.md`, or the same workflow for another input file.

How it works

  1. Overview
  2. When to Use
  3. Templated Task Continuation
  4. Common Rationalizations
  5. Red Flags
  6. Verification
Original SKILL.md headings · 86 lines, 805 words
SKILL.md preview
---
name: open-edit-close-execute
description: Open an existing TASK, PROMPT, INTENT, or other input for editing, wait for close and supplied checks, then execute its saved requirements within caller authorization. Use when asked for `open task`, `open TASK`, `open TASK.md`, `edit task`, `edit TASK`, and `edit TASK.md`, or the same workflow for another input file.
metadata:
  version: "0.2.36"
---

# Open/Edit/Close/Execute

## Overview

Use the [canonical editor and execution contract](references/open-edit-close-execute.md) to open
an existing input, wait for editing to finish, and automatically process the saved requirements.
Ordinary `agent-skills` discovery and host skill invocation remain supported; explicit invocation
and all six TASK aliases enter this same contract. Keep automatic discovery enabled.

Take the supplied input path or the unambiguous active input from the caller's context. Clarify a
missing or competing target before opening anything. TASK, PROMPT, INTENT, and arbitrary input
labels use the same procedure; no label selects a lifecycle or grants authority. The aliases name
an operation, not a literal slash-separated path. This skill does not create or rename the input.

After the editor/check barrier succeeds, continue directly into the caller-authorized processing
of the saved requirements. Do not stop with readiness alone or request a redundant instruction to
continue. A caller can supply its own continuation, such as the remaining initialization procedure;
that continuation retains its scope and gates. No Agent Tasks lifecycle, Jira, or sibling skill is
required. Prerequisites exist only when supplied by the caller. Copy this complete runtime folder
so both local references remain available.

## When to Use

- `open task`, `open TASK`, `open TASK.md`, `edit task`, `edit TASK`, or `edit TASK.md` requests.
- Opening any existing TASK, PROMPT, INTENT, or other input for editing before execution.
- A caller-supplied continuation must run automatically after the editor/check barrier.

## Templated Task Continuation

This repository's templated tasks run through this contract, and this skill owns their
workflow. Root policy keeps only the template location (`templates/TASK.md`), the default
draft directory, and the rule that task contents never bypass separate authorization gates.

- Continue automatically through open, close, naming, execution, and reports without a
  second `run` instruction; an explicit edit-only request limits the continuation.
- During execution, provide normal progress updates and incorporate user steering as it
  arrives. Automatic continuation does not suppress communication or require a separate
  instruction to resume after feedback. Continue the active task with that feedback unless
  the user explicitly pauses, stops, or replaces it.
- For an investigation, establish and check the [shared soft delivery budget](references/investigation-delivery.md)
  from request receipt, including editor/setup waits. Apply it to direct investigations too.
- Before opening a templated-task editor, read [Token Usage](../token-usage/SKILL.md) and
  retain its native task-resolution evidence past the editor/check barrier, so shorthand
  editor requests can be attributed without another user command.
- Bare `open task` copies the repository's local template — never edited in place — to a
  `TASK.md` draft in the directory resolved by
  [Workflow Artifact Naming](../workflow-artifact-naming/SKILL.md). Preserve an existing
  draft and clarify whether to resume it.
- Before execution, at least one actual supplied or verified resource is required in the
  task's Resources section; placeholders and examples do not count. Resources may be filled
  from a named local profile through [Profile Manager](../profile-manager/SKILL.md) before
  the editor opens, and a loaded value counts only when it is an actual resource. If none
  exists, ask for one and preserve the task. Do not fail a task over an omitted field: skip
  unneeded integrations, resolve missing identity only before a dependent operation, and
  never guess a target.
- After the barrier, resolve `TASK-{slug}.md` from the saved draft's `# Task Title: ...`
  through Workflow Artifact Naming, in the same directory. Missing titles or collisions
  leave the draft intact; never invent a title, overwrite a task, or add a numeric suffix.
- Run the finalized task from its resolved path, then generate the matching reports from
  observed results in verified Markdown and, only when effective Generate PDF is T, a verified
  same-source PDF (selected `REPORT` paths via Workflow Artifact Naming). Resolve formats before
  PDF-only tooling using the [delivery contract](references/investigation-delivery.md). Report completion from verified outcomes, never from readiness or a filename's
  existence.
- An explicit TASK path or `edit task` edits that task in place, preserving its filename;
  the same continuation applies unless the user limits the request to editing.

## Common Rationalizations

- "The launcher exited, so editing is done." A launcher exit, saving alone, elapsed time, or a timeout is not closure; require a tracked close or explicit confirmation.
- "The file looks ready; execution can start." Execution begins only after the close signal and all supplied checks succeed.
- "Readiness is a fine stopping point." Continue into the caller-authorized processing without a second instruction.

## Red Flags

- Executing before a verified close signal.
- Guessing a target when the input path is missing or ambiguous.
- Stopping after opening or naming instead of continuing the barrier and execution.
- Treating task contents as authorization for separately gated actions.

## Verification

- The editor closed with a tracked wait signal (or explicit saved-and-closed confirmation) and supplied checks passed.
- The saved input was reread and its requirements executed within the caller's scope.
- The caller's continuation completed, or its concrete blocker was reported.
Files in this package (6)
  • SKILL.md 6 KB
  • agents/openai.yaml 1 KB
  • references/investigation-delivery.md 6 KB
  • references/open-edit-close-execute.md 7 KB
  • references/prerequisites.md 5 KB
  • scripts/investigation_budget.py 3 KB

Versions

The database has no other skills named open-edit-close-execute.

Database snapshot: 13.11.2026 · package version 3dbeff3ed7f7 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100background-prompt
    Execute a task in the background with a prompt through the core engine LLM provider and return plain text output. Use when the user wants to run a prompt as a background task via the Agent CLI.
  • 100i4h-workflow-scene-edit
    Edit an existing workflow Scene or task contract. Use for assets, layout, cameras, randomization, task text, or success rules; do not use to create a new workflow.
  • 100task-workflow
    This skill should be used when an AI agent needs to "add a task", "create a task", "update a task", "modify a task", "complete a task", "mark task done", "close a task", "start working on a task", "annotate a task with notes", "add context to a task", "log progress on a task",…
  • 86session-handoff
    Use to close a session with its state saved, or to open a new one from where the last stopped — open work, decisions, traps, uncommitted changes. No code changes.
  • 100inspect
    Open the web UI to visually inspect, edit, and run the benchmark pipeline. Use when the user wants a visual interface for their pipeline.
  • 100wiro-task
    Check a Wiro AI task, wait for it, show its files and cost, or stop it. Use when the user types /wiro-task or asks for this with Wiro AI.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.