2btwist/dotagents/oneshot

Claude Code and agent skill by 2btwist/dotagents

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Direct execution for small, contained, low-risk tasks (rename, color change, one-line fix) — skip full Research→Plan→Implement, with safety checks.

Install

npx vskill@latest install 2btwist/dotagents/oneshot

SKILL.md

The same SKILL.md was published earlier as 2btwist/dotagents/oneshot.

One-shot

The escape hatch for changes too small for Research → Plan → Implement: renames, visual tweaks (color, spacing, font size), one-line fixes with an obvious cause, copy/config changes, a missing import/type/prop. Contained, obvious, low-risk.

Escalate the moment it isn't (grill it, or /plan if you were told to plan): new features, changes touching multiple concerns, unclear root cause (use /research first), anything that cascades (shared utilities, widely-consumed types), or any "while I'm here" temptation.

Execution style

Execute immediately — no preamble, start with the first grep/read/edit. Pause only if a bail-out condition trips.

Steps

  1. Locate. Grep/Glob/LS to the exact file(s). Three searches is the budget: if the target is still not pinned to a file:line after the third, STOP — not one-shot material; offer /research, or a grill to settle it.
  2. Read fully (no limit/offset) the target and any close neighbour (import, type definition).
  3. Safety check before editing. Grep the symbol repo-wide, and answer all three:
  • How many other call sites are there? More than two → STOP → grill it.
  • Does a test pin this behaviour? Yes → the test is the spec; if the fix changes what it asserts, STOP → grill it.
  • Is there an invariant here the change could break (a null guard, an ordering assumption, a cached value)? Yes → STOP → grill it.

Three clean answers is the only way past this step.

  1. Edit. Minimal diff. No unrelated cleanup, no drive-by refactor.
  2. Verify. Run the smallest relevant check (type check / lint / targeted test, or the project's single check command). One attempt to fix a failure it surfaces; a second failure means the cause was not what you thought — STOP → grill it.
  3. Report. What changed (file:line), what you verified, and anything you noticed but deliberately left untouched.

Guardrails

  • No scope creep — editing a third file means escalate.
  • No half-measures — if it can't be done cleanly in one shot, escalate; don't ship a partial fix.
  • No commits without asking.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
5

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/2btwist/dotagents/oneshot/badge)](https://verified-skill.com/skills/2btwist/dotagents/oneshot)

Something wrong with this skill or its page? Tell us.