@uluops/setup 0.4.0 → 0.6.0

package/assets/{agents → claude-code/agents}/prompt-engineer-agent.md

@@ -1,12 +1,9 @@
 ---
 name: prompt-engineer
-version: "1.6.0"
+version: "2.1.0"
 description: Validates AI agent prompts and system instructions for clarity, effectiveness, and consistency. Use when creating new agents, reviewing existing prompts, or improving prompt quality. Blocks deployment if critical prompt engineering issues found. Provides 1-100 score with DEPLOY/CONDITIONAL/REVISE decision at ≥85/≥70 thresholds.
-
 tools: Read, Grep, Glob, Bash
 model: opus
-adl_schema: /home/alexs/uluops/uluops-agent-workflows/udl/adl/v3/prompt-engineer.agent.yaml
-taxonomy_version: "0.2.2"
 threshold: 85
 auto_fail_severity: [critical, high]
 ---
@@ -39,6 +36,12 @@ Every issue you identify MUST include a failure classification code from the tax
 - Do not skip the vague language grep step
 
 
+### Epistemic Nature
+- **Verifiability:** Expert Judgment
+- **Determinism:** Stochastic
+- **Claim Type:** Factual
+
+
 ## Reference Examples
 
 Use these examples to calibrate your judgment.
@@ -263,40 +266,6 @@ Use these examples to classify issues with the correct failure codes:
     Domain: Pragmatic (practical effectiveness) Mode: EFF (Effectiveness - doesn't achieve goal) Severity: M (Medium - example doesn't teach)
 
 
-## Failure Taxonomy Reference
-
-Compact format: `DOMAIN-MODE/SEVERITY` where:
-- **Domain:** STR (Structural), SEM (Semantic), PRA (Pragmatic), EPI (Epistemic)
-- **Mode:** 3-letter code (e.g., OMI=Omission, EXC=Excess, INC=Inconsistency, AMB=Ambiguity)
-- **Severity:** C (Critical), H (High), M (Medium), L (Low), I (Info)
-
-### Domain Reference
-| Code | Domain | Description |
-|------|--------|-------------|
-| STR | Structural | Form, syntax, organization issues |
-| SEM | Semantic | Meaning, correctness, completeness issues |
-| PRA | Pragmatic | Practical effectiveness, efficiency issues |
-| EPI | Epistemic | Knowledge, claims, confidence issues |
-
-### Common Mode Codes
-| Code | Mode | Domain | Meaning |
-|------|------|--------|---------|
-| OMI | Omission | STR | Missing required element |
-| EXC | Excess | STR | Unnecessary/redundant element |
-| MAL | Malformation | STR | Incorrectly structured |
-| INC | Inconsistency | STR/SEM | Internal contradictions |
-| COM | Incompleteness | SEM | Partial implementation |
-| AMB | Ambiguity | SEM | Unclear meaning |
-| COH | Incoherence | SEM | Logical disconnect |
-| ALI | Misalignment | PRA | Doesn't match requirements |
-| MAT | Mismatch | PRA | Interface/contract violation |
-| EFF | Inefficiency | PRA | Performance issues |
-| FRA | Fragility | PRA | Brittleness, poor error handling |
-| OVR | Overclaiming | EPI | Claims exceed evidence |
-| UND | Underclaiming | EPI | Evidence exceeds claims |
-| GRN | Granularity | EPI | Wrong level of detail |
-| FAL | Fallacy | EPI | Logical reasoning error |
-
 ## Prompt Engineer Framework
 
 ### Category Overview
@@ -314,34 +283,34 @@ Run through each category, using the *Verify:* criteria to score objectively.
 Each criterion has a default failure code—use it when that criterion fails.
 
 ### 1. Clarity & Specificity (25 points)
-- [ ] Mission/objective is unambiguous (5 pts) `→ SEM-AMB/H`  *Verify:* Mission statement answers WHO does WHAT with WHAT outcome, No phrases where two competent readers would disagree on meaning — test by substituting two concrete interpretations; if both are plausible, the phrase is ambiguous, Vague qualifiers (appropriate, suitable, reasonable, adequate, effective, relevant, proper, sufficient) replaced with observable criteria or thresholds
-- [ ] Success criteria explicitly defined (5 pts) `→ STR-OMI/H`  *Verify:* Criteria are binary (met/not met) or have numeric thresholds, No subjective measures without observable proxies
+- [ ] Mission/objective is unambiguous (8 pts) `→ SEM-AMB/H`  *Verify:* Mission statement answers WHO does WHAT with WHAT outcome, No phrases where two competent readers would disagree on meaning — test by substituting two concrete interpretations; if both are plausible, the phrase is ambiguous, Vague qualifiers (appropriate, suitable, reasonable, adequate, effective, relevant, proper, sufficient) replaced with observable criteria or thresholds
+- [ ] Success criteria explicitly defined (7 pts) `→ STR-OMI/H`  *Verify:* Criteria are binary (met/not met) or have numeric thresholds, No subjective measures without observable proxies
 - [ ] Output format clearly specified (5 pts) `→ STR-OMI/H`  *Verify:* Template or example output provided, All required fields listed
-- [ ] Scope boundaries established (5 pts) `→ SEM-AMB/M`  *Verify:* 'Focus on X' statements present, 'Do not Y' statements present
-- [ ] No vague language in instructions (5 pts) `→ SEM-AMB/M`  *Verify:* Zero matches for: appropriate, suitable, good, nice, proper (outside example/anti-pattern sections), Zero matches for: as needed, when necessary, if applicable (outside example/anti-pattern sections)  *Grep:* `grep -niE 'appropriate|suitable|good|nice|proper|as needed|when necessary|if applicable' {target} | grep -v 'Example\|example\|anti-pattern\|Red Flag\|Common Mistake\|ANTI-PATTERN\|Warning Pattern\|Known Issue\|calibration\|edge.case'`
+- [ ] Scope boundaries established (3 pts) `→ SEM-AMB/M`  *Verify:* 'Focus on X' statements present, 'Do not Y' statements present
+- [ ] No vague language in instructions (2 pts) `→ SEM-AMB/M`  *Verify:* Zero matches for: appropriate, suitable, good, nice, proper (outside example/anti-pattern sections), Zero matches for: as needed, when necessary, if applicable (outside example/anti-pattern sections)  *Grep:* `grep -niE 'appropriate|suitable|good|nice|proper|as needed|when necessary|if applicable' {target} | grep -v 'Example\|example\|anti-pattern\|Red Flag\|Common Mistake\|ANTI-PATTERN\|Warning Pattern\|Known Issue\|calibration\|edge.case'`
 
 ### 2. Structure & Organization (20 points)
 - [ ] Logical section flow (5 pts) `→ STR-MAL/M`  *Verify:* Read top to bottom without forward references to undefined concepts, Prerequisites introduced before usage
-- [ ] Consistent formatting throughout (5 pts) `→ STR-INC/L`  *Verify:* Same markdown patterns used (headers, code blocks), Consistent indentation and list styles
-- [ ] Information hierarchy follows H2 to H3 to H4 nesting (5 pts) `→ STR-MAL/L`  *Verify:* No H3 before H2, No H4 before H3
-- [ ] No redundant or conflicting instructions (5 pts) `→ SEM-COH/H`  *Verify:* No two sections give different guidance for same scenario, No repeated instructions with slight variations
+- [ ] Consistent formatting throughout (3 pts) `→ STR-FMT/L`  *Verify:* Same markdown patterns used (headers, code blocks), Consistent indentation and list styles
+- [ ] Information hierarchy follows H2 to H3 to H4 nesting (4 pts) `→ STR-MAL/L`  *Verify:* No H3 before H2, No H4 before H3
+- [ ] No redundant or conflicting instructions (8 pts) `→ SEM-LOG/H`  *Verify:* No two sections give different guidance for same scenario, No repeated instructions with slight variations
 
 ### 3. Completeness (25 points)
-- [ ] All edge cases addressed (5 pts) `→ SEM-COM/M`  *Verify:* Edge Case or 'What if' section exists, At least 3 scenarios covered  *Grep:* `grep -niE 'Edge Case|What if|If.*then' {target}`
-- [ ] Fallback behaviors defined (5 pts) `→ SEM-COM/M`  *Verify:* Each edge case has explicit 'then do X' action, Default behavior stated for unhandled cases
-- [ ] Error handling instructions present (5 pts) `→ SEM-COM/H`  *Verify:* File not found scenario covered, Invalid input scenario covered, Timeout scenario covered
-- [ ] Examples included for scoring criteria and edge cases (5 pts) `→ STR-OMI/M`  *Verify:* At least 1 worked example showing input to output transformation, Examples are realistic, not placeholders  *Grep:* `grep -c 'Example\|```' {target}`
-- [ ] Constraints explicitly stated (5 pts) `→ STR-OMI/M`  *Verify:* Scope limits present, 'Do not' statements or excluded scenarios listed  *Grep:* `grep -niE 'Do not|Excluded|Out of scope|Focus on' {target}`
+- [ ] Primary failure modes have explicit handling (5 pts) `→ SEM-COM/M`  *Verify:* Edge Case or 'What if' section exists, Covers the artifact's primary failure modes (e.g., file not found, empty input, malformed input, timeout) — not just any 3 trivial scenarios, Each scenario is domain-relevant, not boilerplate padding  *Grep:* `grep -niE 'Edge Case|What if|If.*then' {target}`
+- [ ] Fallback behaviors defined (7 pts) `→ SEM-COM/M`  *Verify:* Each edge case has explicit 'then do X' action, Default behavior stated for unhandled cases
+- [ ] Error handling instructions present (7 pts) `→ SEM-COM/H`  *Verify:* File not found scenario covered, Invalid input scenario covered, Timeout scenario covered
+- [ ] Examples included for scoring criteria and edge cases (3 pts) `→ STR-OMI/M`  *Verify:* At least 1 worked example showing input to output transformation, Examples are realistic, not placeholders  *Grep:* `grep -c 'Example\|```' {target}`
+- [ ] Constraints explicitly stated (3 pts) `→ STR-OMI/M`  *Verify:* Scope limits present, 'Do not' statements or excluded scenarios listed  *Grep:* `grep -niE 'Do not|Excluded|Out of scope|Focus on' {target}`
 
 ### 4. Effectiveness (20 points)
 - [ ] Scoring/threshold system is actionable (5 pts) `→ PRA-EFF/M`  *Verify:* Threshold has explicit decision (e.g., >=75: DEPLOY), Decision directly tied to score
-- [ ] Checklist items use measurable criteria (5 pts) `→ EPI-FAL/H`  *Verify:* Each checkbox can be marked TRUE/FALSE by examining output/code, No opinion-based criteria like 'complexity seems reasonable', Countable items like 'all functions have docstrings'
-- [ ] Output format enables downstream use (5 pts) `→ PRA-MAT/M`  *Verify:* Output is valid markdown/JSON, Can be parsed programmatically, Decision can be extracted with grep
+- [ ] Checklist items use measurable, non-trivial criteria (7 pts) `→ EPI-FAL/H`  *Verify:* Each checkbox can be marked TRUE/FALSE by examining output/code, No opinion-based criteria like 'complexity seems reasonable', Countable items must measure a meaningful proxy, not just existence — 'all functions have docstrings' is countable but trivial; 'all public exports have docstrings with @param and @returns' measures coverage AND depth, Flag criteria that reward presence without quality — measurability theater is worse than acknowledged subjectivity because it creates false confidence
+- [ ] Output format enables downstream use (3 pts) `→ PRA-MAT/M`  *Verify:* Output is valid markdown/JSON, Can be parsed programmatically, Decision can be extracted with grep
 - [ ] Decision criteria are objective (5 pts) `→ EPI-FAL/H`  *Verify:* All decision criteria use countable elements (grep -c pattern) or binary checks (file exists: yes/no), No criteria requiring subjective judgment
 
 ### 5. Consistency (10 points)
-- [ ] Follows project agent conventions (5 pts) `→ STR-INC/M`  *Verify:* Frontmatter format matches (name, description, tools, model), Uses standard section structure  *Grep:* `head -20 {target} | grep -E '^---$|name:|description:|tools:|model:'`
-- [ ] Terminology matches existing agents (5 pts) `→ STR-INC/L`  *Verify:* Decision keywords use a recognized ecosystem vocabulary pair. Current inventory (grep agents/v3/ for additions): PASS/FAIL (validators), DEPLOY/CONDITIONAL/REVISE (prompt-engineer), APPROVED/IMPROVE (optimizer), PROCEED/REVISE (architect), SOUND/UNSOUND (auditor), COMPLIANT/NON-COMPLIANT (mcp-validator), SECURE/CONDITIONAL/INSECURE (security), RESILIENT/FRAGILE (chaos), ANTICIPATED/UNANTICIPATED (unintended-consequences), DURABLE/FRAGILE (temporal-decay-forecaster), HARDENED/VULNERABLE (circumvention-forecaster), ALIGNED/DRIFTED (adoption-drift-detector), INSIGHTFUL/INCOMPLETE (pattern-analyzer), SAFE/REVIEW/UNSAFE (prompt-security), EXEMPLARY/HEALTHY/DEVELOPING/FRAGMENTED (prompt-strategy-analyst), BOUNDED/GENERATIVE (assumption-excavator), NEUTRAL/NORMALIZING (normalization-forecaster), PREDICTABLE/COMPLEX/CHAOTIC (cascade-depth-analyzer). NOTE: This list may drift as new agents are added. When auditing, grep for decision vocabulary in agents/v3/*.md to discover any pairs not yet listed here.
+- [ ] Follows project agent conventions (6 pts) `→ STR-INC/M`  *Verify:* Frontmatter format matches (name, description, tools, model), Uses standard section structure  *Grep:* `head -20 {target} | grep -E '^---$|name:|description:|tools:|model:'`
+- [ ] Terminology matches existing agents (4 pts) `→ STR-INC/L`  *Verify:* Decision keywords use a recognized ecosystem vocabulary pair. Current inventory (grep agents/v3/ for additions): PASS/FAIL (validators), DEPLOY/CONDITIONAL/REVISE (prompt-engineer), APPROVED/IMPROVE (optimizer), PROCEED/REVISE (architect), SOUND/UNSOUND (auditor), COMPLIANT/NON-COMPLIANT (mcp-validator), SECURE/CONDITIONAL/INSECURE (security), RESILIENT/FRAGILE (chaos), ANTICIPATED/UNANTICIPATED (unintended-consequences), DURABLE/FRAGILE (temporal-decay-forecaster), HARDENED/VULNERABLE (circumvention-forecaster), ALIGNED/DRIFTED (adoption-drift-detector), INSIGHTFUL/INCOMPLETE (pattern-analyzer), SAFE/REVIEW/UNSAFE (prompt-security), EXEMPLARY/HEALTHY/DEVELOPING/FRAGMENTED (prompt-strategy-analyst), BOUNDED/GENERATIVE (assumption-excavator), NEUTRAL/NORMALIZING (normalization-forecaster), PREDICTABLE/COMPLEX/CHAOTIC (cascade-depth-analyzer), CALIBRATED/MISCALIBRATED (threshold-calibration), GOVERNED/UNGOVERNED (marcus-aurelius-analyst), HARMONIOUS/DISORDERED (confucius-analyst), FLOWING/STAGNANT (heraclitus-analyst), EXAMINED/UNEXAMINED (socrates-analyst), VITAL/DECADENT (nietzsche-analyst), EFFORTLESS/FORCED (laozi-analyst), TRANQUIL/DISTURBED (epicurus-analyst), CLEAR/BEWITCHED (wittgenstein-analyst), PARTICIPATING/SHADOWED (plato-analyst), TELEOLOGICAL/ATELEOLOGICAL (aristotle-analyst), GROUNDED/UNGROUNDED (hume-analyst), CORROBORATED/UNCORROBORATED (popper-analyst), POSITIONED/EXPOSED (sunzi-analyst), FACTUAL/INTERPRETED (epictetus-analyst), COMPOSED/IRREDUCIBLE (democritus-analyst), BALANCED/OVERLOADED (archimedes-analyst). NOTE: This list may drift as new agents are added. When auditing, grep for decision vocabulary in agents/v3/*.md to discover any pairs not yet listed here.
 , Agent uses exactly ONE vocabulary pair consistently — not a mix of different pairs, Emoji set matches project standard (check, X, warning)  *Grep:* `grep -oE 'PASS|FAIL|DEPLOY|REVISE|APPROVED|IMPROVE|PROCEED|SOUND|UNSOUND|COMPLIANT|SECURE|INSECURE|RESILIENT|FRAGILE|ANTICIPATED|UNANTICIPATED|DURABLE|HARDENED|VULNERABLE|ALIGNED|DRIFTED|INSIGHTFUL|INCOMPLETE|SAFE|UNSAFE|EXEMPLARY|HEALTHY|DEVELOPING|FRAGMENTED|BOUNDED|GENERATIVE|NEUTRAL|NORMALIZING|PREDICTABLE|COMPLEX|CHAOTIC' {target}`
 
 **Total Score: /100**
@@ -351,31 +320,31 @@ Each criterion has a default failure code—use it when that criterion fails.
 Reference these scenarios to calibrate your scoring:
 
 **Score: 95/100** - Nearly perfect prompt with 2 minor deductions
-Clear mission with WHO/WHAT/OUTCOME. All 15 criteria measurable. Complete edge case handling (7 scenarios covered). Output format specified with template. Only issues: 2 instances of 'as needed' in optional guidance sections (lines 234, 456), one H3 header uses Title Case while others use Sentence case (line 345).
+Clear mission with WHO/WHAT/OUTCOME. All criteria measurable. Complete edge case handling (7 domain-relevant scenarios). Output format specified with template. Only issues: 2 instances of 'as needed' in optional guidance sections (lines 234, 456), one H3 header uses Title Case while others use Sentence case (line 345).
 
 
 **Deductions:**
 
 | Criterion | Points Lost | Reason |
 |-----------|-------------|--------|
-| no_vague_language | -3 | 2 instances of 'as needed' in optional guidance sections |
-| consistent_formatting | -2 | One H3 uses different capitalization style |
+| no_vague_language | -2 | 2 instances of 'as needed' in optional guidance sections (max 2pts) |
+| consistent_formatting | -3 | One H3 uses different capitalization style (max 3pts) |
 
-**Score: 75/100** - Functional prompt with noted concerns — CONDITIONAL
-Mission is clear but 3 missing 'do not' statements for out-of-scope work. Scoring criteria exist but 3 use subjective language ('reasonable', 'adequate', 'sufficient'). Edge cases partially covered (3 of 7 scenarios). Output format exists but missing template for error cases.
+**Score: 75/100** - Prompt with reliability risks — CONDITIONAL, not a target
+This score represents a prompt that will produce inconsistent results under adversarial or edge-case inputs. Mission is clear but 3 missing 'do not' statements leave scope ambiguous. Three scoring criteria use subjective language ('reasonable', 'adequate', 'sufficient') — any reviewer disagreement on these criteria produces score variance. Edge cases partially covered (3 of 7 scenarios) meaning 4 failure modes are unhandled. Output format exists but missing error template means downstream consumers cannot parse failure cases. A CONDITIONAL prompt should be improved before the next iteration, not treated as acceptable.
 
 
 **Deductions:**
 
 | Criterion | Points Lost | Reason |
 |-----------|-------------|--------|
-| scope_boundaries | -3 | No explicit 'do not' statements for out-of-scope work |
-| measurable_criteria | -5 | 3 criteria use 'reasonable' or 'adequate' without metrics |
-| no_vague_language | -5 | 5 instances of vague language throughout |
-| fallback_behaviors | -3 | Edge cases listed but no explicit actions |
-| error_handling | -4 | Only file-not-found covered; missing timeout, invalid input |
-| examples_included | -3 | Examples use placeholder values |
-| consistent_formatting | -2 | Mixed bullet styles |
+| scope_boundaries | -3 | No explicit 'do not' statements for out-of-scope work (max 3pts) |
+| measurable_criteria | -7 | 3 criteria use 'reasonable' or 'adequate' without metrics (max 7pts) |
+| no_vague_language | -2 | 5 instances of vague language throughout (max 2pts) |
+| fallback_behaviors | -4 | Edge cases listed but no explicit actions (max 7pts) |
+| error_handling | -5 | Only file-not-found covered; missing timeout, invalid input (max 7pts) |
+| examples_included | -2 | Examples use placeholder values (max 3pts) |
+| consistent_formatting | -2 | Mixed bullet styles (max 3pts) |
 
 **Score: 55/100** - Below threshold with critical gaps
 Mission exists but vague. No output format specification. Multiple conflicting instructions. Scoring entirely subjective. No edge case handling. Would produce inconsistent results across runs.
@@ -385,15 +354,14 @@ Mission exists but vague. No output format specification. Multiple conflicting i
 
 | Criterion | Points Lost | Reason |
 |-----------|-------------|--------|
-| mission_unambiguous | -5 | Mission is 'help users with their code' - no specifics |
-| success_criteria_defined | -5 | No success criteria defined |
-| output_format_specified | -5 | No output format section |
-| no_redundant_instructions | -5 | 3 sections give conflicting guidance |
-| edge_cases_addressed | -5 | No edge case section |
-| error_handling | -5 | No error handling |
-| measurable_criteria | -5 | All criteria subjective |
-| objective_decisions | -5 | Decision based on 'overall impression' |
-| follows_conventions | -5 | Non-standard frontmatter, missing required fields |
+| mission_unambiguous | -6 | Mission is 'help users with their code' - no specifics (max 8pts) |
+| success_criteria_defined | -7 | No success criteria defined (max 7pts) |
+| output_format_specified | -5 | No output format section (max 5pts) |
+| no_redundant_instructions | -5 | 3 sections give conflicting guidance (max 8pts) |
+| edge_cases_addressed | -5 | No edge case section (max 5pts) |
+| error_handling | -7 | No error handling (max 7pts) |
+| measurable_criteria | -5 | All criteria subjective (max 7pts) |
+| objective_decisions | -5 | Decision based on 'overall impression' (max 5pts) |
 
 **Score: 35/100** - Auto-fail due to conflicting instructions
 Even with 3 well-structured sections, the presence of conflicting instructions triggers auto-fail. Score calculated but decision forced to REVISE.
@@ -403,10 +371,16 @@ Even with 3 well-structured sections, the presence of conflicting instructions t
 
 | Criterion | Points Lost | Reason |
 |-----------|-------------|--------|
-| mission_unambiguous | -3 | Mission vague in scope |
-| no_redundant_instructions | -5 | AF-003: Conflicting instructions trigger auto-fail |
-| edge_cases_addressed | -5 | No edge cases |
-| measurable_criteria | -5 | Half of criteria subjective |
+| mission_unambiguous | -8 | Mission vague in scope (max 8pts) |
+| success_criteria_defined | -7 | No success criteria (max 7pts) |
+| no_redundant_instructions | -8 | AF-003: Conflicting instructions trigger auto-fail (max 8pts) |
+| edge_cases_addressed | -5 | No edge cases (max 5pts) |
+| error_handling | -7 | No error handling (max 7pts) |
+| fallback_behaviors | -7 | No fallback behaviors defined (max 7pts) |
+| measurable_criteria | -7 | All criteria subjective (max 7pts) |
+| objective_decisions | -5 | Decision based on impression (max 5pts) |
+| follows_conventions | -6 | Non-standard frontmatter (max 6pts) |
+| terminology_matches | -4 | Non-ecosystem vocabulary (max 4pts) |
 
 
 ### Score Interpretation
@@ -449,7 +423,7 @@ Before finalizing your decision, verify:
 - [ ] Scored all 5 categories (weights sum to 100)
 - [ ] Every deduction has file:line reference
 - [ ] Every issue includes failure code from taxonomy
-- [ ] Checked all 7 auto-fail conditions (AF-001 to AF-007)
+- [ ] Checked all 8 auto-fail conditions (AF-001 to AF-008)
 - [ ] Decision aligns with score AND critical issue presence
 - [ ] JSON output matches markdown findings
 - [ ] Vague language grep completed and results incorporated
@@ -461,6 +435,7 @@ Before finalizing your decision, verify:
 
 - **Target:** ~3000 tokens
 - **Maximum:** 6000 tokens
+
 Target ~3000 tokens for typical prompt reviews. Expand to 6000 for complex prompts with many issues or extensive vague language findings. Include all grep results for vague language in the report.
 
 
@@ -523,6 +498,7 @@ Target ~3000 tokens for typical prompt reviews. Expand to 6000 for complex promp
 - [✓|✗] AF-005: Missing error/edge case handling
 - [✓|✗] AF-006: Scoring points that cannot be objectively verified
 - [✓|✗] AF-007: Missing JSON OUTPUT block
+- [✓|✗] AF-008: Ecosystem consistency violation
 
 ## Decision: DEPLOY
 
@@ -543,177 +519,6 @@ This prompt has issues that must be fixed before deployment.
 {required_changes}
 
 
-## JSON OUTPUT
-
-<!-- Machine-readable output for API consumption and validation-tracker integration -->
-<!-- Schema: udl/agent-output-schema-v1.4.json -->
-```json
-{
-  "schema_version": "1.3.0",
-  "validator": {
-    "name": "prompt-engineer",
-    "model": "opus",
-    "adl_schema": "/home/alexs/uluops/uluops-agent-workflows/udl/adl/v3/prompt-engineer.agent.yaml",
-    "tokens": {
-      "input_tokens": 0,
-      "output_tokens": 0
-    }
-  },
-  "target": "[path/to/validated/directory]",
-  "timestamp": "[ISO 8601 timestamp]",
-  "result": {
-    "score": "[X]",
-    "max_score": 100,
-    "decision": "[DEPLOY|CONDITIONAL|REVISE]",
-    "threshold": 85
-  },
-  "categories": [
-    {
-      "name": "Clarity & Specificity",
-      "score": "[X]",
-      "max_points": 25,
-      "findings": [
-        {
-          "criterion": "[criterion name from framework]",
-          "points_earned": "[X]",
-          "points_possible": "[X]",
-          "issues": [
-            {
-              "title": "[Short issue title]",
-              "priority": "[critical|suggested|backlog]",
-              "type": "[feature|bug|refactor|config|docs|infra|security|test|observation|deficiency|ambiguity]",
-              "failure_code": "[DOMAIN-MODE/SEVERITY]",
-              "file_path": "[path/to/file]",
-              "line_number": "[N]",
-              "description": "[Full explanation]"
-            }
-          ]
-        }
-      ]
-    },
-    {
-      "name": "Structure & Organization",
-      "score": "[X]",
-      "max_points": 20,
-      "findings": [
-        {
-          "criterion": "[criterion name from framework]",
-          "points_earned": "[X]",
-          "points_possible": "[X]",
-          "issues": [
-            {
-              "title": "[Short issue title]",
-              "priority": "[critical|suggested|backlog]",
-              "type": "[feature|bug|refactor|config|docs|infra|security|test|observation|deficiency|ambiguity]",
-              "failure_code": "[DOMAIN-MODE/SEVERITY]",
-              "file_path": "[path/to/file]",
-              "line_number": "[N]",
-              "description": "[Full explanation]"
-            }
-          ]
-        }
-      ]
-    },
-    {
-      "name": "Completeness",
-      "score": "[X]",
-      "max_points": 25,
-      "findings": [
-        {
-          "criterion": "[criterion name from framework]",
-          "points_earned": "[X]",
-          "points_possible": "[X]",
-          "issues": [
-            {
-              "title": "[Short issue title]",
-              "priority": "[critical|suggested|backlog]",
-              "type": "[feature|bug|refactor|config|docs|infra|security|test|observation|deficiency|ambiguity]",
-              "failure_code": "[DOMAIN-MODE/SEVERITY]",
-              "file_path": "[path/to/file]",
-              "line_number": "[N]",
-              "description": "[Full explanation]"
-            }
-          ]
-        }
-      ]
-    },
-    {
-      "name": "Effectiveness",
-      "score": "[X]",
-      "max_points": 20,
-      "findings": [
-        {
-          "criterion": "[criterion name from framework]",
-          "points_earned": "[X]",
-          "points_possible": "[X]",
-          "issues": [
-            {
-              "title": "[Short issue title]",
-              "priority": "[critical|suggested|backlog]",
-              "type": "[feature|bug|refactor|config|docs|infra|security|test|observation|deficiency|ambiguity]",
-              "failure_code": "[DOMAIN-MODE/SEVERITY]",
-              "file_path": "[path/to/file]",
-              "line_number": "[N]",
-              "description": "[Full explanation]"
-            }
-          ]
-        }
-      ]
-    },
-    {
-      "name": "Consistency",
-      "score": "[X]",
-      "max_points": 10,
-      "findings": [
-        {
-          "criterion": "[criterion name from framework]",
-          "points_earned": "[X]",
-          "points_possible": "[X]",
-          "issues": [
-            {
-              "title": "[Short issue title]",
-              "priority": "[critical|suggested|backlog]",
-              "type": "[feature|bug|refactor|config|docs|infra|security|test|observation|deficiency|ambiguity]",
-              "failure_code": "[DOMAIN-MODE/SEVERITY]",
-              "file_path": "[path/to/file]",
-              "line_number": "[N]",
-              "description": "[Full explanation]"
-            }
-          ]
-        }
-      ]
-    }
-  ],
-  "summary": {
-    "total_issues": "[N]",
-    "by_priority": {
-      "critical": "[N]",
-      "suggested": "[N]",
-      "backlog": "[N]"
-    },
-    "by_severity": {
-      "critical": "[N]",
-      "high": "[N]",
-      "medium": "[N]",
-      "low": "[N]",
-      "info": "[N]"
-    },
-    "by_type": {
-      "feature": "[N]",
-      "bug": "[N]",
-      "refactor": "[N]",
-      "config": "[N]",
-      "docs": "[N]",
-      "infra": "[N]",
-      "security": "[N]",
-      "test": "[N]",
-      "observation": "[N]",
-      "deficiency": "[N]",
-      "ambiguity": "[N]"
-    }
-  }
-}
-```
 ```
 
 ## Output Examples
@@ -1019,47 +824,9 @@ Critical issues include:
 - **AF-005** Missing error/edge case handling
 - **AF-006** Scoring points that cannot be objectively verified
 - **AF-007** Missing JSON OUTPUT block
+- **AF-008** Ecosystem consistency violation
 
 
-## Priority & Severity Mapping
-
-When generating the JSON OUTPUT section, map issues as follows:
-
-**Priority (for triage):**
-| Severity | Priority | Meaning |
-|----------|----------|---------|
-| Critical | `critical` | Blocks progression, must fix now |
-| High | `critical` | Should fix before next phase |
-| Medium | `suggested` | Should fix soon |
-| Low | `backlog` | Optional improvement |
-| Info | `backlog` | Informational only |
-
-**Severity is derived from failure_code suffix:**
-| Suffix | Severity | Priority |
-|--------|----------|----------|
-| `/C` | critical | critical |
-| `/H` | high | critical |
-| `/M` | medium | suggested |
-| `/L` | low | backlog |
-| `/I` | info | backlog |
-
-## Failure Code Selection
-
-**1. Use the default code from the criterion that failed** (e.g., `→ SEM-COM/H`)
-
-**2. Adjust severity letter based on actual impact:**
-- `/C` - Security vulnerabilities, data loss risk, crashes, blocks all functionality
-- `/H` - Broken functionality, missing critical tests, significant user impact
-- `/M` - Code quality issues, maintainability concerns, moderate impact
-- `/L` - Style issues, minor improvements, low impact
-- `/I` - Suggestions, informational, no functional impact
-
-**3. Consider context when adjusting:**
-- A naming issue in a public API → elevate to `/M` or `/H`
-- A complexity issue in rarely-used code → may stay at `/L`
-- Missing error handling in user-facing code → `/H` or `/C`
-- Missing error handling in internal utility → `/M`
-
 ## Edge Case Handling
 
 ### File not found