cohemm/prism

Multi-perspective AI team analysis for better decisions

5Stars on the repository
20Mods indexed here, across every type
3mo agoLast push, which is what freshness is scored on
MITLicence, which decides whether bodies are shown

analyze

01

cohemm/prism

Skill Claude CodeCodex

Runs multi-perspective analysis via MCP server orchestration. Thin wrapper that collects user input then delegates all processing to prismanalyze MCP tool. Ontology scope is automatically resolved from brownfield default repositories. General-purpose analysis engine — any topic can be analyzed.

5 3mo ago A 54 tokens original MIT

brownfield

02

cohemm/prism

Skill Claude CodeCodex

Scan and manage brownfield repository defaults for interviews.

5 3mo ago A 12 tokens original MIT

incident

03

cohemm/prism

Skill Claude CodeCodex

Incident root cause analysis with UX impact — takes an incident description (text + optional screenshots) as input, runs multi-perspective analysis including UX impact perspective via prismanalyze MCP pipeline, producing a developer-facing RCA report. Use this skill for "incident analysis", "incident postmortem"…

5 3mo ago A 106 tokens original MIT

prd

04

cohemm/prism

Skill Claude CodeCodex

PRD policy conflict analysis — takes a prd.md file as input, runs multi-perspective analysis against codebase policy documents (ontology), and generates a PM-friendly report. Use this skill for "prd analysis", "PRD policy review", "review PRD", "check PRD conflicts", "PRD policy check", or any request mentioning a…

5 3mo ago A 112 tokens original MIT

setup

05

cohemm/prism

Skill Claude CodeCodex

Scan brownfield repositories and configure defaults for prism skills.

5 3mo ago A 12 tokens original MIT

test-analyze

06

cohemm/prism

Skill Claude CodeCodex

Live test runner for prism:analyze skill. Executes full multi-agent workflow against real codebases, verifies JSON producer/consumer contracts between phases, checks real prisminterview verification scores, and validates that findings without code evidence receive low scores. Use this skill when the user says "test…

5 3mo ago A 96 tokens original MIT