From 3f5a1a236ab279c362a7f1b02e8ec790695b049a Mon Sep 17 00:00:00 2001 From: vp Date: Mon, 30 Mar 2026 01:11:23 +0300 Subject: [PATCH 1/6] feat: add ontology-and-logic-audit skill --- README.md | 3 +- ontology-and-logic-audit/SKILL.md | 130 ++++++++++++++++++++++++++++++ 2 files changed, 132 insertions(+), 1 deletion(-) create mode 100644 ontology-and-logic-audit/SKILL.md diff --git a/README.md b/README.md index c4e96bb..0ad3d48 100644 --- a/README.md +++ b/README.md @@ -13,7 +13,7 @@ Install: place these skill folders under `$CODEX_HOME/skills` ## Skills -This repo currently includes 16 skills: +This repo currently includes 17 skills: | Skill | Folder | Description | | --- | --- | --- | @@ -22,6 +22,7 @@ This repo currently includes 16 skills: | iOS Debugger Agent | `ios-debugger-agent` | Uses XcodeBuildMCP to build, launch, and debug the current iOS app on a booted simulator, including UI inspection, interaction, screenshots, and log capture. | | macOS Menubar Tuist App | `macos-menubar-tuist-app` | Builds, refactors, or reviews macOS menubar apps that use Tuist and SwiftUI, with emphasis on manifest ownership, store-layer architecture, and reliable local launch scripts. | | macOS SwiftPM App Packaging (No Xcode) | `macos-spm-app-packaging` | Scaffolds, builds, packages, signs, and optionally notarizes SwiftPM-based macOS apps without requiring an Xcode project. | +| Ontology and Logic Audit | `ontology-and-logic-audit` | Analyzes a spec, argument, requirement set, prompt, or conceptual framework by validating ontology before logic, then surfaces hidden assumptions, modality shifts, and minimal repairs. | | Orchestrate Batch Refactor | `orchestrate-batch-refactor` | Plans and executes larger refactor or rewrite efforts with dependency-aware parallel analysis and implementation using clearly scoped work packets. | | Project Skill Audit | `project-skill-audit` | Analyzes a project's past Codex sessions, memory, existing local skills, and conventions to recommend the highest-value new skills or updates to existing ones. | | React Component Performance | `react-component-performance` | Diagnoses slow React components by finding re-render churn, expensive render work, unstable props, and list bottlenecks, then suggests targeted optimizations and validation steps. | diff --git a/ontology-and-logic-audit/SKILL.md b/ontology-and-logic-audit/SKILL.md new file mode 100644 index 0000000..89130e3 --- /dev/null +++ b/ontology-and-logic-audit/SKILL.md @@ -0,0 +1,130 @@ +--- +name: ontology-and-logic-audit +description: Analyze a spec, argument, requirement set, prompt, or conceptual framework by validating ontology before logic. Use when the user asks for rigorous critique, category-mistake detection, hidden-assumption analysis, modality separation, or minimal repair of a draft whose core concepts may be confused. +--- + +# Ontology and Logic Audit + +Audit a draft by checking whether its subject matter is being described under the right kind of thing before evaluating whether the reasoning is valid. Use this skill for specs, prompts, arguments, requirements, architecture write-ups, and conceptual frameworks where the user wants a stricter critique than a normal review. + +## Core Principles + +- Define the terms before judging the argument. +- Test ontology before logic: if the subject is misclassified, say so immediately. +- Do not rescue a broken ontology by charitable interpretation. +- Distinguish established claims from hypotheses. +- Keep hypotheses testable and repairs minimal. + +## Workflow + +### 1) Normalize the terms + +- Extract the key terms the draft relies on. +- Define each term in plain working language before analyzing the reasoning. +- If the draft uses the same term in more than one sense, flag the shift and state the normalized meaning you will use. + +Typical examples: + +- thing vs property +- capability vs permission +- rule vs observation +- implementation detail vs architectural boundary +- empirical claim vs conceptual claim + +### 2) Validate the ontology first + +- Ask what kind of thing each central claim is about. +- Check whether the draft treats one category as if it were another. +- Look for category mistakes, collapsed distinctions, or a framework that conflicts with obvious real cases. + +If the ontology fails: + +- say so immediately +- give one concrete counterexample +- label the failure as `categorical` or `empirical` +- stop trying to save the original framing through reinterpretation + +### 3) Audit the logic + +Only after the ontology is sound enough to proceed: + +- surface hidden assumptions +- test whether the conclusion follows from the premises +- check whether the draft changes modality mid-argument +- identify contradictions, unsupported leaps, or circular reasoning +- check for salvage by trivialization, where a strong claim only survives after being weakened into a tautology + +### 4) Separate modalities + +Call out the kinds of claim being made: + +- conceptual necessity +- empirical possibility +- practical constraint +- policy requirement +- contingent implementation choice + +If the draft slides from one modality to another, name the shift explicitly. + +### 5) Present the audit as an argument + +Structure the response in this order: + +1. normalized terms +2. ontology verdict +3. premises +4. reasoning steps +5. conclusion +6. minimal repair, if needed + +Prefer short, direct statements over rhetorical flourish. + +## Output Expectations + +Return a compact audit that includes: + +1. `Term normalization` + - the key terms and the meanings used for the audit + +2. `Ontology verdict` + - whether the framing is sound + - the exact category mistake or empirical conflict, if present + +3. `Logic findings` + - hidden assumptions + - invalid steps + - contradictions or trivializations + +4. `Modalities` + - the types of necessity/possibility claims being made + +5. `Minimal repair` + - the smallest restatement that preserves the user's goal under a sound ontology + +## Failure Shields + +- Do not skip ontology and jump straight to stylistic or logical critique. +- Do not soften a category mistake into a vague “may need clarification”. +- Do not treat every disagreement as an ontology failure; distinguish conceptual errors from missing evidence. +- Do not inflate hypotheses into established conclusions. +- Do not rewrite the draft wholesale when a smaller repair will do. + +## Good Fits + +Use this skill when the user asks for: + +- rigorous critique of a spec or proposal +- ontology-first review of a prompt or framework +- category-mistake detection +- hidden-assumption analysis +- separation of possibility, necessity, and policy claims +- a minimal repair of a conceptually confused draft + +## Poor Fits + +Prefer a different skill when the user wants: + +- ordinary code review of a diff +- bug root-cause investigation +- broad brainstorming without a concrete draft to audit +- style-only copy editing From b3287a016e54b1b01b138cd54661ad1894be1e30 Mon Sep 17 00:00:00 2001 From: vp Date: Mon, 30 Mar 2026 01:16:11 +0300 Subject: [PATCH 2/6] refine: align ontology skill with source prompt --- README.md | 2 +- ontology-and-logic-audit/SKILL.md | 132 ++---------------------------- 2 files changed, 8 insertions(+), 126 deletions(-) diff --git a/README.md b/README.md index 0ad3d48..d7dbdb7 100644 --- a/README.md +++ b/README.md @@ -22,7 +22,7 @@ This repo currently includes 17 skills: | iOS Debugger Agent | `ios-debugger-agent` | Uses XcodeBuildMCP to build, launch, and debug the current iOS app on a booted simulator, including UI inspection, interaction, screenshots, and log capture. | | macOS Menubar Tuist App | `macos-menubar-tuist-app` | Builds, refactors, or reviews macOS menubar apps that use Tuist and SwiftUI, with emphasis on manifest ownership, store-layer architecture, and reliable local launch scripts. | | macOS SwiftPM App Packaging (No Xcode) | `macos-spm-app-packaging` | Scaffolds, builds, packages, signs, and optionally notarizes SwiftPM-based macOS apps without requiring an Xcode project. | -| Ontology and Logic Audit | `ontology-and-logic-audit` | Analyzes a spec, argument, requirement set, prompt, or conceptual framework by validating ontology before logic, then surfaces hidden assumptions, modality shifts, and minimal repairs. | +| Ontology and Logic Audit | `ontology-and-logic-audit` | Applies a Greek-scholastic reasoning pass: define terms, validate ontology before logic, separate modalities, and propose minimal repair when the ontology fails. | | Orchestrate Batch Refactor | `orchestrate-batch-refactor` | Plans and executes larger refactor or rewrite efforts with dependency-aware parallel analysis and implementation using clearly scoped work packets. | | Project Skill Audit | `project-skill-audit` | Analyzes a project's past Codex sessions, memory, existing local skills, and conventions to recommend the highest-value new skills or updates to existing ones. | | React Component Performance | `react-component-performance` | Diagnoses slow React components by finding re-render churn, expensive render work, unstable props, and list bottlenecks, then suggests targeted optimizations and validation steps. | diff --git a/ontology-and-logic-audit/SKILL.md b/ontology-and-logic-audit/SKILL.md index 89130e3..01734c8 100644 --- a/ontology-and-logic-audit/SKILL.md +++ b/ontology-and-logic-audit/SKILL.md @@ -1,130 +1,12 @@ --- name: ontology-and-logic-audit -description: Analyze a spec, argument, requirement set, prompt, or conceptual framework by validating ontology before logic. Use when the user asks for rigorous critique, category-mistake detection, hidden-assumption analysis, modality separation, or minimal repair of a draft whose core concepts may be confused. +description: Critique text, specs, arguments, reviews, requirements, prompts, and conceptual frameworks by defining terms, validating ontology before logic, separating modalities, and proposing minimal repair when the ontology fails. --- -# Ontology and Logic Audit +You are a reasoning assistant grounded in structured inquiry and Greek-scholastic traditions. When responding: -Audit a draft by checking whether its subject matter is being described under the right kind of thing before evaluating whether the reasoning is valid. Use this skill for specs, prompts, arguments, requirements, architecture write-ups, and conceptual frameworks where the user wants a stricter critique than a normal review. - -## Core Principles - -- Define the terms before judging the argument. -- Test ontology before logic: if the subject is misclassified, say so immediately. -- Do not rescue a broken ontology by charitable interpretation. -- Distinguish established claims from hypotheses. -- Keep hypotheses testable and repairs minimal. - -## Workflow - -### 1) Normalize the terms - -- Extract the key terms the draft relies on. -- Define each term in plain working language before analyzing the reasoning. -- If the draft uses the same term in more than one sense, flag the shift and state the normalized meaning you will use. - -Typical examples: - -- thing vs property -- capability vs permission -- rule vs observation -- implementation detail vs architectural boundary -- empirical claim vs conceptual claim - -### 2) Validate the ontology first - -- Ask what kind of thing each central claim is about. -- Check whether the draft treats one category as if it were another. -- Look for category mistakes, collapsed distinctions, or a framework that conflicts with obvious real cases. - -If the ontology fails: - -- say so immediately -- give one concrete counterexample -- label the failure as `categorical` or `empirical` -- stop trying to save the original framing through reinterpretation - -### 3) Audit the logic - -Only after the ontology is sound enough to proceed: - -- surface hidden assumptions -- test whether the conclusion follows from the premises -- check whether the draft changes modality mid-argument -- identify contradictions, unsupported leaps, or circular reasoning -- check for salvage by trivialization, where a strong claim only survives after being weakened into a tautology - -### 4) Separate modalities - -Call out the kinds of claim being made: - -- conceptual necessity -- empirical possibility -- practical constraint -- policy requirement -- contingent implementation choice - -If the draft slides from one modality to another, name the shift explicitly. - -### 5) Present the audit as an argument - -Structure the response in this order: - -1. normalized terms -2. ontology verdict -3. premises -4. reasoning steps -5. conclusion -6. minimal repair, if needed - -Prefer short, direct statements over rhetorical flourish. - -## Output Expectations - -Return a compact audit that includes: - -1. `Term normalization` - - the key terms and the meanings used for the audit - -2. `Ontology verdict` - - whether the framing is sound - - the exact category mistake or empirical conflict, if present - -3. `Logic findings` - - hidden assumptions - - invalid steps - - contradictions or trivializations - -4. `Modalities` - - the types of necessity/possibility claims being made - -5. `Minimal repair` - - the smallest restatement that preserves the user's goal under a sound ontology - -## Failure Shields - -- Do not skip ontology and jump straight to stylistic or logical critique. -- Do not soften a category mistake into a vague “may need clarification”. -- Do not treat every disagreement as an ontology failure; distinguish conceptual errors from missing evidence. -- Do not inflate hypotheses into established conclusions. -- Do not rewrite the draft wholesale when a smaller repair will do. - -## Good Fits - -Use this skill when the user asks for: - -- rigorous critique of a spec or proposal -- ontology-first review of a prompt or framework -- category-mistake detection -- hidden-assumption analysis -- separation of possibility, necessity, and policy claims -- a minimal repair of a conceptually confused draft - -## Poor Fits - -Prefer a different skill when the user wants: - -- ordinary code review of a diff -- bug root-cause investigation -- broad brainstorming without a concrete draft to audit -- style-only copy editing +1. Define key terms (scholastic style) to remove ambiguity; if the author uses them inconsistently, flag it and state your normalization. +2. Validate ontology first: test whether the framework collapses the subject via a category mistake or conflict with real examples. If it does, say so immediately, give a concrete counterexample, label the failure (categorical vs empirical), and do not rescue it by charitable interpretation. +3. Analyze the logic: surface hidden assumptions; check for inconsistencies and for “salvage by trivialization” (saving the argument only by reducing it to a tautology). State this explicitly when it occurs. +4. Infer and separate modalities in the text (kinds of possibility and necessity). +5. Present a structured argument (premises → steps → conclusion); distinguish hypotheses from established claims, and keep hypotheses testable. If the ontology fails, propose the minimal repair or restate the problem under a sound ontology and, where feasible, re-run the argument. From 74c5808c499211e1f264e7b6064e0544f1cf3a39 Mon Sep 17 00:00:00 2001 From: vp Date: Mon, 30 Mar 2026 01:24:03 +0300 Subject: [PATCH 3/6] refine: update ontology skill descriptions --- README.md | 2 +- ontology-and-logic-audit/SKILL.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index d7dbdb7..4796c40 100644 --- a/README.md +++ b/README.md @@ -22,7 +22,7 @@ This repo currently includes 17 skills: | iOS Debugger Agent | `ios-debugger-agent` | Uses XcodeBuildMCP to build, launch, and debug the current iOS app on a booted simulator, including UI inspection, interaction, screenshots, and log capture. | | macOS Menubar Tuist App | `macos-menubar-tuist-app` | Builds, refactors, or reviews macOS menubar apps that use Tuist and SwiftUI, with emphasis on manifest ownership, store-layer architecture, and reliable local launch scripts. | | macOS SwiftPM App Packaging (No Xcode) | `macos-spm-app-packaging` | Scaffolds, builds, packages, signs, and optionally notarizes SwiftPM-based macOS apps without requiring an Xcode project. | -| Ontology and Logic Audit | `ontology-and-logic-audit` | Applies a Greek-scholastic reasoning pass: define terms, validate ontology before logic, separate modalities, and propose minimal repair when the ontology fails. | +| Ontology and Logic Audit | `ontology-and-logic-audit` | Applies a Greek-scholastic reasoning pass: define terms, validate ontology before logic, separate modalities, and propose minimal repair when the ontology or logic fails. | | Orchestrate Batch Refactor | `orchestrate-batch-refactor` | Plans and executes larger refactor or rewrite efforts with dependency-aware parallel analysis and implementation using clearly scoped work packets. | | Project Skill Audit | `project-skill-audit` | Analyzes a project's past Codex sessions, memory, existing local skills, and conventions to recommend the highest-value new skills or updates to existing ones. | | React Component Performance | `react-component-performance` | Diagnoses slow React components by finding re-render churn, expensive render work, unstable props, and list bottlenecks, then suggests targeted optimizations and validation steps. | diff --git a/ontology-and-logic-audit/SKILL.md b/ontology-and-logic-audit/SKILL.md index 01734c8..e8a7e7b 100644 --- a/ontology-and-logic-audit/SKILL.md +++ b/ontology-and-logic-audit/SKILL.md @@ -1,6 +1,6 @@ --- name: ontology-and-logic-audit -description: Critique text, specs, arguments, reviews, requirements, prompts, and conceptual frameworks by defining terms, validating ontology before logic, separating modalities, and proposing minimal repair when the ontology fails. +description: Critique text, specs, arguments, reviews, requirements, prompts, and conceptual frameworks by defining terms, validating ontology before logic, separating modalities, and proposing minimal repair when the ontology or logic fails. Applies a Greek-scholastic reasoning pass: define terms, validate ontology before logic, separate modalities. --- You are a reasoning assistant grounded in structured inquiry and Greek-scholastic traditions. When responding: From 066123f66d992451091825008f299b2c396ef857 Mon Sep 17 00:00:00 2001 From: vp Date: Mon, 30 Mar 2026 01:38:30 +0300 Subject: [PATCH 4/6] refine: align ontology skill copy --- README.md | 2 +- ontology-and-logic-audit/SKILL.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 4796c40..8e9b6a9 100644 --- a/README.md +++ b/README.md @@ -22,7 +22,7 @@ This repo currently includes 17 skills: | iOS Debugger Agent | `ios-debugger-agent` | Uses XcodeBuildMCP to build, launch, and debug the current iOS app on a booted simulator, including UI inspection, interaction, screenshots, and log capture. | | macOS Menubar Tuist App | `macos-menubar-tuist-app` | Builds, refactors, or reviews macOS menubar apps that use Tuist and SwiftUI, with emphasis on manifest ownership, store-layer architecture, and reliable local launch scripts. | | macOS SwiftPM App Packaging (No Xcode) | `macos-spm-app-packaging` | Scaffolds, builds, packages, signs, and optionally notarizes SwiftPM-based macOS apps without requiring an Xcode project. | -| Ontology and Logic Audit | `ontology-and-logic-audit` | Applies a Greek-scholastic reasoning pass: define terms, validate ontology before logic, separate modalities, and propose minimal repair when the ontology or logic fails. | +| Ontology and Logic Audit | `ontology-and-logic-audit` | Audit and critique specs, arguments, reviews, requirements, prompts, and conceptual frameworks using a Greek-scholastic reasoning pass: define key terms, validate ontology before logic, separate modalities, and propose minimal repair when ontology fails or the reasoning is invalid. | | Orchestrate Batch Refactor | `orchestrate-batch-refactor` | Plans and executes larger refactor or rewrite efforts with dependency-aware parallel analysis and implementation using clearly scoped work packets. | | Project Skill Audit | `project-skill-audit` | Analyzes a project's past Codex sessions, memory, existing local skills, and conventions to recommend the highest-value new skills or updates to existing ones. | | React Component Performance | `react-component-performance` | Diagnoses slow React components by finding re-render churn, expensive render work, unstable props, and list bottlenecks, then suggests targeted optimizations and validation steps. | diff --git a/ontology-and-logic-audit/SKILL.md b/ontology-and-logic-audit/SKILL.md index e8a7e7b..d27526f 100644 --- a/ontology-and-logic-audit/SKILL.md +++ b/ontology-and-logic-audit/SKILL.md @@ -1,6 +1,6 @@ --- name: ontology-and-logic-audit -description: Critique text, specs, arguments, reviews, requirements, prompts, and conceptual frameworks by defining terms, validating ontology before logic, separating modalities, and proposing minimal repair when the ontology or logic fails. Applies a Greek-scholastic reasoning pass: define terms, validate ontology before logic, separate modalities. +description: Audit and critique specs, arguments, reviews, requirements, prompts, and conceptual frameworks using a Greek-scholastic reasoning pass: define key terms, validate ontology before logic, separate modalities, and propose minimal repair when ontology fails or the reasoning is invalid. --- You are a reasoning assistant grounded in structured inquiry and Greek-scholastic traditions. When responding: From fc87bbcec5d20dbd94f1a7f6be3ac2ba182e7a0b Mon Sep 17 00:00:00 2001 From: vp Date: Mon, 30 Mar 2026 01:45:51 +0300 Subject: [PATCH 5/6] fix: quote ontology skill description yaml --- ontology-and-logic-audit/SKILL.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ontology-and-logic-audit/SKILL.md b/ontology-and-logic-audit/SKILL.md index d27526f..19c2d5d 100644 --- a/ontology-and-logic-audit/SKILL.md +++ b/ontology-and-logic-audit/SKILL.md @@ -1,6 +1,6 @@ --- name: ontology-and-logic-audit -description: Audit and critique specs, arguments, reviews, requirements, prompts, and conceptual frameworks using a Greek-scholastic reasoning pass: define key terms, validate ontology before logic, separate modalities, and propose minimal repair when ontology fails or the reasoning is invalid. +description: "Audit and critique specs, arguments, reviews, requirements, prompts, and conceptual frameworks using a Greek-scholastic reasoning pass: define key terms, validate ontology before logic, separate modalities, and propose minimal repair when ontology fails or the reasoning is invalid." --- You are a reasoning assistant grounded in structured inquiry and Greek-scholastic traditions. When responding: From 9c94d53aff1dde0ba63163b34e58266ed844b28d Mon Sep 17 00:00:00 2001 From: vp Date: Mon, 30 Mar 2026 01:48:46 +0300 Subject: [PATCH 6/6] feat: add openai agent config for ontology skill --- ontology-and-logic-audit/agents/openai.yaml | 4 ++++ 1 file changed, 4 insertions(+) create mode 100644 ontology-and-logic-audit/agents/openai.yaml diff --git a/ontology-and-logic-audit/agents/openai.yaml b/ontology-and-logic-audit/agents/openai.yaml new file mode 100644 index 0000000..449e95a --- /dev/null +++ b/ontology-and-logic-audit/agents/openai.yaml @@ -0,0 +1,4 @@ +interface: + display_name: "Ontology and Logic Audit" + short_description: "Audit ontology before evaluating logic" + default_prompt: "Use $ontology-and-logic-audit to audit this spec, argument, prompt, or framework by defining key terms, validating ontology before logic, separating modalities, and proposing minimal repair."