All skills
simota avatar

/magi

@35ffd55
by shingo imotasimota/agent-skills85 stars
15

Deliberating decisions and founder priorities through multi-perspective, named-expert, and YC-style advisory lenses. Use for verdicts, office hours, or expert critique; not implementation.

Use this Skill: https://skilld.dev/gh/simota/agent-skills/magi

This session only. Nothing lands on disk.

referencereframing-toolkit.md

≈841 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Reframing Toolkit (absorbed from Refract)

3-axis perspective rotation methodology for enriching Magi deliberations. Previously a standalone agent (Refract), now integrated as a Magi pre-deliberation tool.


When to Apply

Use before Magi deliberation when:

  • The problem feels "stuck" in one frame
  • Logos/Pathos/Sophia all evaluate from the same assumptions
  • Blind spots may exist that the 3 perspectives won't catch alone
  • Complex strategic decisions with multiple stakeholders

3 Rotation Axes

Axis Name Question Transforms
Scope Field / Scope What scale are you viewing this at? Zoom Out · Zoom In · Cross-Domain · Temporal Zoom
Standpoint Standpoint / Vantage Whose position are you viewing this from? Role Shift · Hierarchy Shift · Timeline Shift · Adversarial Shift
Lens Angle / Lens What frame are you viewing this through? Systems Lens · Inversion Lens · First Principles · Narrative Lens

Quick Reframing Process

Step 1: DECODE — Identify the current frame

  • What is the dominant metaphor? ("cost problem" / "tech problem" / "people problem")
  • What assumptions are implicit?
  • Whose perspective dominates?

Step 2: ROTATE — Apply at least one transform per axis

Field rotation examples:

  • Zoom Out: "How does this look at the organization level?"
  • Zoom In: "What happens to one specific user?"
  • Cross-Domain: "How does [unrelated industry] solve similar problems?"

Standpoint rotation examples:

  • Role Shift: "How does the intern / CEO / customer see this?"
  • Adversarial: "If I were the competitor, how would I exploit this?"
  • Timeline: "How would we judge this decision in 5 years?"

Lens rotation examples:

  • Systems: "What feedback loops exist? Where are leverage points?"
  • Inversion: "What if we do the exact opposite?"
  • First Principles: "Strip away assumptions — what remains?"

Step 3: ILLUMINATE — Surface what was hidden

  • Per rotation: "What's newly visible?" + "What was hiding?"
  • Flag any cognitive biases detected

Step 4: Feed into Magi Deliberation

  • Provide reframed perspectives as additional input to Logos/Pathos/Sophia
  • Each perspective considers the blind spots identified

Integration with Magi Deliberation

Problem → Reframing (this toolkit) → Enriched problem statement
                                           ↓
                                    Magi Deliberation
                                    ├── Logos (+ blind spots)
                                    ├── Pathos (+ stakeholder views)
                                    └── Sophia (+ strategic reframes)
                                           ↓
                                    Verdict (better informed)

Common Cognitive Biases to Check

Bias Sign Reframing Counter
Anchoring First option dominates Inversion lens
Sunk cost "We've already invested..." Timeline shift (future-only view)
Availability Recent events overweighted Cross-domain rotation
Confirmation Only supporting evidence cited Adversarial standpoint
Status quo "It's always been this way" First principles lens
Groupthink No dissent in discussion Role shift to outsider

Source: SKILL.md on GitHub

1 alert13d5 checks · Risk HIGH
  • Gen Agent Trust Hub13d

    The skill facilitates complex decision-making but introduces significant security risks through its 'Engine Mode' and 'multi' recipe. These modes execute external CLI tools (codex, agy) using shell commands that incorporate user-provided context. Most notably, the skill explicitly instructs the agent to use a flag that bypasses permission checks (--dangerously-skip-permissions) when invoking the Antigravity CLI. This creates a high-risk vector for unauthorized command execution and bypasses platform security controls.

  • Socket13d

    No alerts

  • Snyk13d

    Risk: LOW · No issues

  • Runlayer6mo

    2/7 files flagged

  • ZeroLeaks5mo

    Score: 93/100 · 2 sections analyzed

Signed by skilld at 35ffd55. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 2 days ago.

Activeupdated 2 weeks ago

README badge

README badge for simota/agent-skills/magi