radarist/structured-analytic-skills / skills/analysis-of-competing-hypotheses
analysis-of-competing-hypotheses
VerifiedRanks rival explanations of a question by which one the evidence contradicts least (Heuer's Analysis of Competing Hypotheses — ACH — CIA), then names the single piece of evidence the ranking hinges on. Use when a question has two or more plausible explanations and confirmation bias would be costly — "why is adoption stalling?", "is this signal real or just noise?", "which of these causes explains the drop?", or before a headline conclusion ships. Not for factual single-answer questions (use `grounded-answer`) or when only one hypothesis exists.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 11.2 KB |
| agents/openai.yaml | config | 377 B |
| evals/evals.json | config | 4.2 KB |
| examples/adoption-stall.json | config | 1.3 KB |
| scripts/ach.py | script | 15.7 KB |
evals/evals.json
View on GitHub{
"skill_name": "analysis-of-competing-hypotheses",
"evals": [
{
"id": 1,
"case_id": "ach-pos-1",
"kind": "positive",
"prompt": "Our enterprise product's adoption stalled last quarter and I can think of at least three explanations — slower sales cycle, a competitor's launch, or budget freezes. Help me work out which explanation the evidence actually supports before I brief the board.",
"expected_output": "A response that satisfies these checks: Enumerates 3+ hypotheses including a null/noise hypothesis; Builds an evidence-by-hypothesis matrix with C/N/I/NA marks; Ranks hypotheses by fewest inconsistencies, not most consistencies; Names the single most diagnostic piece of evidence and a sensitivity statement; Lists leftover uncertainty and what would disconfirm the runner-up.",
"files": [],
"expectations": [
"Enumerates 3+ hypotheses including a null/noise hypothesis",
"Builds an evidence-by-hypothesis matrix with C/N/I/NA marks",
"Ranks hypotheses by fewest inconsistencies, not most consistencies",
"Names the single most diagnostic piece of evidence and a sensitivity statement",
"Lists leftover uncertainty and what would disconfirm the runner-up"
],
"skills": [
"analysis-of-competing-hypotheses"
]
},
{
"id": 2,
"case_id": "ach-pos-2",
"kind": "positive",
"prompt": "Is the sudden jump in GitHub stars for this open-source model a real signal of adoption or just noise from a viral post? There are several competing explanations and I want a disciplined comparison, not a gut call.",
"expected_output": "A response that satisfies these checks: Treats 'noise' as an explicit hypothesis; Grades each observation against every hypothesis independently; Reports the leading hypothesis with its inconsistency count and the pivotal evidence.",
"files": [],
"expectations": [
"Treats 'noise' as an explicit hypothesis",
"Grades each observation against every hypothesis independently",
"Reports the leading hypothesis with its inconsistency count and the pivotal evidence"
],
"skills": [
"analysis-of-competing-hypotheses"
]
},
{
"id": 3,
"case_id": "ach-pos-3",
"kind": "positive",
"prompt": "Run an ACH on why our competitor suddenly cut prices 30% — pivot, distress, or market-share grab?",
"expected_output": "A response that satisfies these checks: Runs the eight-step ACH procedure; Includes a sensitivity check on the load-bearing evidence; Uses the output template with ranking, most-diagnostic evidence, sensitivity, loose ends.",
"files": [],
"expectations": [
"Runs the eight-step ACH procedure",
"Includes a sensitivity check on the load-bearing evidence",
"Uses the output template with ranking, most-diagnostic evidence, sensitivity, loose ends"
],
"skills": [
"analysis-of-competing-hypotheses"
]
},
{
"id": 4,
"case_id": "ach-neg-1",
"kind": "negative",
"prompt": "What year did NVIDIA release the H100 GPU and what was its launch price? Give me a sourced answer.",
"expected_output": "A response that satisfies these checks: Does not build a hypothesis matrix; Answers the factual question with a verified source (grounded-answer behaviour).",
"files": [],
"expectations": [
"Does not build a hypothesis matrix",
"Answers the factual question with a verified source (grounded-answer behaviour)"
],
"skills": [
"grounded-answer"
]
},
{
"id": 5,
"case_id": "ach-neg-2",
"kind": "negative",
"prompt": "I already believe there is a 60% chance the supplier misses the deadline; a new report says their factory is at 70% capacity. Update my probability.",
"expected_output": "A response that satisfies these checks: Performs a Bayesian update of a single belief rather than enumerating and ranking rival hypotheses.",
"files": [],
"expectations": [
"Performs a Bayesian update of a single belief rather than enumerating and ranking rival hypotheses"
],
"skills": [
"bayesian-update"
]
}
]
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |