github-copilot-superpowers task-direction-approval.instructions.md

Instructions that require user approval before changing the requested technical approach, architecture, library, or delivery method. They also define how to explain failures and present alternatives.

In plain words
What is it for?
Use them when an implementation fails and you are considering another library, framework, architecture, manual process, or result than the one originally requested.
Why use it?
They prevent an agent from silently replacing the user’s chosen direction with a workaround or a materially different solution.

Instructions file for GitHub Copilot

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/kasuken/github-copilot-superpowers/task-direction-approval
Clone the repo
git clone --depth 1 https://github.com/kasuken/github-copilot-superpowers

Made for: GitHub Copilot.

Per session 247 This file is loaded in full into every session.
When invoked 247 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00247 $0.00247
Opus 5 $0.00123 $0.00123
Sonnet 5 $0.00049 $0.00049
Haiku 4.5 $0.00025 $0.00025

Measured 2d ago against content hash aa9c17284ac5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

github-copilot-superpowers task-direction-approval.instructions.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.github/instructions/task-direction-approval.instructions.md · 45 lines

What it actually says

Task Direction Approval Guidelines

When User Approval is Required

NEVER do without asking:

  • Switch from original tech/library to alternatives
  • Replace automated approach with manual workarounds
  • Change architecture or design patterns
  • Deliver different results than requested

ALWAYS do before changing direction:

  1. Analyze failure root cause clearly
  2. Explain situation to user transparently
  3. Present multiple solution options
  4. Wait for user's explicit choice

Communication Protocol

See the task-direction-approval skill for reusable communication templates and examples.

Exception Cases

OK to change immediately:

  • Fix obvious typos or syntax errors
  • Correct clear configuration mistakes
  • User explicitly says "any method is fine"

Must ask user first:

  • Switch libraries/frameworks
  • Change architectural patterns
  • Impact performance/security
  • Affect maintainability

Failure Response Rules

See the investigation-mode skill for the repeated-failure stop rule and investigation workflow.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 45 lines · 247 tokens per session scan A aa9c17284ac5

Subscribe to this mod's changes

github-copilot-superpowers task-direction-approval.instructions.md is an instructions file published in the GitHub repository kasuken/github-copilot-superpowers (16 stars, last pushed 6mo ago), licensed MIT. It adds 247 tokens to every session, about $0.0012 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.