← ClaudeAtlas

god-modelisted

Route broad or high-ambiguity requests through the smallest useful sequence of existing skills. Use when the user asks for god mode, all skills, maximum capability, end-to-end autonomous work, or a task that spans implementation, review, architecture, safety, and handoff. NOT for simple single-skill requests, literal claims of universal expertise, bypassing permissions, or skipping verification.
alinafe82/cognitive-deadlift · ★ 0 · AI & Automation · score 70
Install: claude install-skill alinafe82/cognitive-deadlift
# God Mode ## Purpose Coordinate complex work by selecting, sequencing, and verifying the relevant existing skills instead of pretending one generic skill can do every job. ## Preserves Skill routing judgment, scope control, and operational discipline under broad requests. ## Required Evidence - User goal, constraints, and requested autonomy level. - Available skill names, descriptions, or local skill paths checked for this session. - Risk level and the evidence required by that risk. - Selected skill sequence and at least one skill intentionally skipped. - Verification commands, source files, or review artifacts used before confidence. ## Failure Signs - The assistant says it used every skill without listing what was available. - The response skips a narrower skill that clearly applies. - The assistant claims access to skills, tools, documents, or permissions it has not verified. - The plan maximizes activity instead of choosing the smallest useful sequence. - The final answer hides failed checks, missing evidence, or unresolved risk. ## When To Use - The user explicitly asks for "god mode," "all skills," "maximum capability," or similar broad orchestration. - A task crosses multiple phases such as framing, code reading, tests, implementation, diff review, security, and handoff. - The assistant needs to decide whether several skills should run in sequence. - A broad request risks becoming autopilot unless skill routing and evidence are explicit. ## When Not To Use