Normalize agent workflow temp paths to /tmp/gh-aw/agent

.github/copilot/instructions/workflow-performance.md

@@ -6,13 +6,13 @@ This guide helps you optimize agentic workflow execution efficiency, focusing on
 
 ```bash
 # Analyze recent workflow runs
-gh aw logs workflow-name -c 10
+gh aw logs workflow-name -c 10 -o /tmp/gh-aw/agent/logs/
 
 # Audit specific run for performance insights
 gh aw audit <run-id>
 
 # Download logs for analysis
-gh aw logs --start-date -1w -o /tmp/gh-aw/perf/
+gh aw logs --start-date -1w -o /tmp/gh-aw/agent/perf/
 ```
 
 ## Performance Targets
@@ -33,8 +33,8 @@ gh aw logs --start-date -1w -o /tmp/gh-aw/perf/
 **Measurement**:
 ```bash
 # Analyze token distribution in logs
-gh aw logs workflow-name -c 1
-grep -i "token" /tmp/gh-aw/logs/workflow-name/*.log
+gh aw logs workflow-name -c 1 -o /tmp/gh-aw/agent/logs/
+grep -R -i "token" /tmp/gh-aw/agent/logs/
 ```
 
 **Optimization Strategy**:
@@ -72,10 +72,10 @@ Use the github tool to get full issue details with all fields.
 **Measurement**:
 ```bash
 # Count tool calls in logs
-grep "tool_use" /tmp/gh-aw/logs/workflow/*.log | wc -l
+grep -R "tool_use" /tmp/gh-aw/agent/logs/ | wc -l
 
 # Identify repeated calls
-grep "tool_use" /tmp/gh-aw/logs/workflow/*.log | sort | uniq -c | sort -rn
+grep -R "tool_use" /tmp/gh-aw/agent/logs/ | sort | uniq -c | sort -rn
 ```
 
 **Optimization Strategy**:
@@ -118,7 +118,7 @@ steps:
   setup-env:
     name: Setup Environment
     run: |
-      mkdir -p /tmp/gh-aw/work
+      mkdir -p /tmp/gh-aw/agent/work
       git config --global user.name "github-actions[bot]"
       git config --global user.email "github-actions[bot]@users.noreply.github.com"
 ```
@@ -243,8 +243,8 @@ steps:
 steps:
   setup-dirs:
     run: |
-      mkdir -p /tmp/gh-aw/work
-      mkdir -p /tmp/gh-aw/output
+      mkdir -p /tmp/gh-aw/agent/work
+      mkdir -p /tmp/gh-aw/agent/output
 ```
 
 **Pre-install Tools** (if needed):

.github/skills/jqschema/SKILL.md

@@ -4,19 +4,13 @@ description: Infer JSON structure and types with jq-based schema discovery.
 tools:
   bash:
     - "jq *"
-    - "/tmp/gh-aw/jqschema.sh"
+    - "./.github/skills/jqschema/jqschema.sh"
     - "git"
-steps:
-  - name: Setup jq utilities directory
-    run: |
-      mkdir -p /tmp/gh-aw
-      cp "$GITHUB_WORKSPACE/.github/skills/jqschema/jqschema.sh" /tmp/gh-aw/jqschema.sh
-      chmod +x /tmp/gh-aw/jqschema.sh
 ---
 
 ## jqschema - JSON Schema Discovery
 
-Use `/tmp/gh-aw/jqschema.sh` to discover complex JSON structure.
+Use `./.github/skills/jqschema/jqschema.sh` directly from the repository skill folder to discover complex JSON structure.
 
 ### Purpose
 
@@ -30,13 +24,13 @@ Generate a compact structural schema (keys + types) from JSON input. Use it when
 
 ```bash
 # Analyze a file
-cat data.json | /tmp/gh-aw/jqschema.sh
+cat data.json | ./.github/skills/jqschema/jqschema.sh
 
 # Analyze command output
-echo '{"name": "test", "count": 42, "items": [{"id": 1}]}' | /tmp/gh-aw/jqschema.sh
+echo '{"name": "test", "count": 42, "items": [{"id": 1}]}' | ./.github/skills/jqschema/jqschema.sh
 
 # Analyze GitHub search results
-gh api search/repositories?q=language:go | /tmp/gh-aw/jqschema.sh
+gh api search/repositories?q=language:go | ./.github/skills/jqschema/jqschema.sh
 ```
 
 ### How It Works
@@ -77,7 +71,7 @@ The script transforms JSON data by:
 ```bash
 # Step 1: Get schema with minimal data (fetch just 1 result)
 # This helps understand the structure before requesting large datasets
-echo '{}' | gh api search/repositories -f q="language:go" -f per_page=1 | /tmp/gh-aw/jqschema.sh
+echo '{}' | gh api search/repositories -f q="language:go" -f per_page=1 | ./.github/skills/jqschema/jqschema.sh
 
 # Output shows the schema:
 # {"incomplete_results":"boolean","items":[{...}],"total_count":"number"}
@@ -95,7 +89,7 @@ When using tools like `search_code`, `search_issues`, or `search_repositories`,
 gh api search/code -f q="jq in:file language:bash" -f per_page=1 > /tmp/sample.json
 
 # Generate schema to understand structure
-cat /tmp/sample.json | /tmp/gh-aw/jqschema.sh
+cat /tmp/sample.json | ./.github/skills/jqschema/jqschema.sh
 
 # Now you know which fields exist and can use them in your analysis
 ```

.github/workflows/ab-testing-advisor.md

@@ -268,7 +268,7 @@ Show the concrete before/after diff.
 - [ ] Add `experiments:` section to frontmatter
 - [ ] Add conditional blocks to workflow prompt body using `{{#if experiments.<name> == "<variant>" }}` (value-comparison form — never use the internal `__GH_AW_EXPERIMENTS__` env-var syntax)
 - [ ] Run `gh aw compile <workflow-name>` to regenerate lock file
-- [ ] Monitor experiment artifact uploaded per run to `/tmp/gh-aw/experiments/state.json`
+- [ ] Monitor experiment artifact uploaded per run to `/tmp/gh-aw/agent/experiments/state.json`
 - [ ] After sufficient runs, analyze variant distribution via workflow run artifacts
 - [ ] Document findings and promote winning variant
 
@@ -307,7 +307,7 @@ After completing the primary quest, include a **second issue** (sub-issue of the
 
 Use the `field-presence-checker` agent with file paths `pkg/workflow/compiler_experiments.go` and `actions/setup/js/pick_experiment.cjs`, and field names `analysis_type`, `tags`, `notify`. Use the returned `present`/`evidence` results when deciding which fields are genuinely absent.
 
-Then review what data is currently captured per experiment run (the artifact uploaded to `/tmp/gh-aw/experiments/state.json`) and consider what would be needed for a complete experiment analytics pipeline.
+Then review what data is currently captured per experiment run (the artifact uploaded to `/tmp/gh-aw/agent/experiments/state.json`) and consider what would be needed for a complete experiment analytics pipeline.
 
 Propose concrete improvements in the following areas: