feat: GIA-71 configure dependabot to linear ticket sync workflow

.github/workflows/ci.yml

@@ -33,7 +33,7 @@ jobs:
           python-version: "3.12"
 
       - name: Run Pre-Commit Hooks in CI
-        uses: pre-commit/action@v3.0.1
+        run: uvx pre-commit run --all-files
 
       - name: Install Dependencies
         run: uv pip install -e ".[test]" || uv pip install -r requirements.txt || true
@@ -63,7 +63,7 @@ jobs:
 
       - name: Upload Semgrep results to GitHub Security
         uses: github/codeql-action/upload-sarif@v3
-        if: always()
+        if: always() && github.event_name != 'pull_request'
         with:
           sarif_file: semgrep-results.sarif
 
@@ -79,7 +79,7 @@ jobs:
 
       # Pinned securely to v0.36.0 commit SHA to mitigate CVE-2026-33634
       - name: Run Trivy File System Scanner
-        uses: aquasecurity/trivy-action@ed142fda32cedf951e7f3d537877df4c9ba8f192 # v0.36.0
+        uses: aquasecurity/trivy-action@ed142fd0673e97e23eac54620cfb913e5ce36c25 # v0.36.0
         with:
           scan-type: 'fs'
           ignore-unfixed: true
@@ -89,6 +89,6 @@ jobs:
 
       - name: Upload Trivy results to GitHub Security
         uses: github/codeql-action/upload-sarif@v3
-        if: always()
+        if: always() && github.event_name != 'pull_request'
         with:
           sarif_file: trivy-results.sarif

.github/workflows/dependabot-linear.yml

@@ -0,0 +1,37 @@
+name: Sync Dependabot PRs to Linear
+on:
+  pull_request:
+    types: [opened]
+
+jobs:
+  create-linear-ticket:
+    if: github.actor == 'dependabot[bot]'
+    runs-on: ubuntu-latest
+    steps:
+      - name: Create Linear Issue
+        env:
+          LINEAR_TOKEN: ${{ secrets.LINEAR_TOKEN }}
+          PR_TITLE: ${{ github.event.pull_request.title }}
+          PR_URL: ${{ github.event.pull_request.html_url }}
+        run: |
+          JSON_PAYLOAD=$(jq -n \
+            --arg title "Dependabot: $PR_TITLE" \
+            --arg desc "GitHub PR: $PR_URL\nOpened by Dependabot." \
+            --arg team "a46f213b-5a47-43bb-bea9-ca254bd3dfe2" \
+            --arg project "329d07ee-9041-4706-9244-bd2dc7dfb982" \
+            '{
+              query: "mutation IssueCreate($input: IssueCreateInput!) { issueCreate(input: $input) { success issue { id number identifier url } } }",
+              variables: {
+                input: {
+                  title: $title,
+                  description: $desc,
+                  teamId: $team,
+                  projectId: $project
+                }
+              }
+            }')
+
+          curl -X POST https://api.linear.app/graphql \
+            -H "Content-Type: application/json" \
+            -H "Authorization: $LINEAR_TOKEN" \
+            -d "$JSON_PAYLOAD"

.gitignore

@@ -18,7 +18,3 @@
 # Python bytecode cache
 __pycache__/
 *.pyc
-
-
-
-

.pre-commit-config.yaml

@@ -1,5 +1,5 @@
 # .pre-commit-config.yaml
-exclude: '^(Machine Learning|Cybersecurity|Papers|Study|Templates|Attachments|Miscellaneous|Inbox|__site|__obsidian|node_modules)/'
+exclude: '^(\.obsidian|Coffee|Machine Learning|Cybersecurity|Miscellaneous|PDFs|Papers|Study|Templates|Attachments|Inbox|__site|__obsidian|node_modules)/'
 
 repos:
   # 1. General Repository Hygiene

.skills/skills/clip-to-vault/clip-to-vault/SKILL.md

@@ -14,6 +14,7 @@ The **clip-to-vault** skill transforms web content into vault notes through synt
 ### Input
 
 The agent accepts a **URL** pointing to:
+
 - Technical blog posts (e.g., Anthropic blog, Papers with Code)
 - Official documentation (LLM APIs, ML frameworks)
 - Conference talks or transcripts
@@ -135,6 +136,7 @@ constitutional_ai, value_alignment, llm_training, ai_safety
 ## Inbox Awareness
 
 When the user requests a fast capture rather than a full synthesis (e.g., "clip this to inbox", "save to inbox quickly"), the skill should create a minimal capture instead:
+
 - **File location**: `Inbox/[title].md` (at vault root)
 - **Minimal format**: Title, URL, 1–2 sentence summary, no internal links or full synthesis required
 - The Inbox serves as a staging area for rapid ideas; items can be processed into full notes later
@@ -149,4 +151,3 @@ The skill can also **process existing Inbox items** into proper notes. When the
 - **Check for duplicates**: Before creating a new note, verify that the vault does not already contain this information.
 - **Style consistency**: Match the vault's academic tone. No casual language, no contractions, third-person perspective.
 - **Link with intention**: Every wikilink should be meaningful. Readers should want to follow them to deepen understanding.
-

.skills/skills/connect-ideas/connect-ideas/SKILL.md

@@ -14,11 +14,13 @@ This skill analyzes two related notes or a note and its linked references to unc
 ### 1. Gather Input Notes
 
 Request the user provide:
+
 - Primary note name or path (the focus note)
 - Secondary note name or path (the comparison note), OR a list of linked notes to analyze
 - Any specific aspect to emphasize (e.g., methodological, theoretical, practical)
 
 Read both notes fully. Document:
+
 - Primary themes and key concepts in each
 - Explicit claims, definitions, and arguments
 - Examples, evidence, or cited resources
@@ -27,6 +29,7 @@ Read both notes fully. Document:
 ### 2. Identify Shared Concepts
 
 Compare the notes and list **shared conceptual terrain**:
+
 - Overlapping terminology or frameworks
 - Related but differently-named concepts
 - Common underlying principles or assumptions
@@ -37,11 +40,13 @@ Create a table or bullet list showing how each note addresses the shared concept
 ### 3. Detect Contradictions or Tensions
 
 Scan for potential disagreements:
+
 - Do the notes reach conflicting conclusions on the same topic?
 - Do they prioritize different aspects (e.g., theory vs. practice)?
 - Are there unstated assumptions in each note that contradict the other?
 
 For each tension found:
+
 - State it clearly and objectively
 - Explain the reasoning behind each position
 - Assess whether this is a true contradiction or a complementary perspective
@@ -50,6 +55,7 @@ For each tension found:
 ### 4. Discover Non-Obvious Connections
 
 Look beyond surface-level similarity:
+
 - Does one note provide background context that explains a gap in the other?
 - Can concepts from Note A apply as a solution or example in Note B?
 - Do the two notes address the same problem from different angles?
@@ -60,18 +66,21 @@ Articulate each hidden connection in a short sentence with clear reasoning.
 ### 5. Suggest Specific Wikilinks
 
 For each connection found, recommend:
+
 - A wikilink to add to the primary note (format: `[[note-name|anchor-text]]`)
 - A reciprocal wikilink to add to the secondary note
 - Optional: A suggested context phrase where the link should appear
 
 Prioritize links that would:
+
 - Help future readers jump between related ideas
 - Fill gaps in the vault's knowledge graph
 - Make tacit connections explicit
 
 ### 6. Optional: Produce a Mermaid Diagram
 
 If the relationship is complex or multi-faceted, create a Mermaid diagram showing:
+
 - The two main notes as nodes
 - Shared concepts as intermediate nodes
 - Arrows labeled with the type of relationship (contradicts, extends, applies-to, provides-context-for, etc.)
@@ -82,7 +91,7 @@ Use a graph or flowchart format. Keep labels concise.
 
 Structure the analysis as follows:
 
-```
+````
 ## Connection Analysis: [Note A] ↔ [Note B]
 
 ### Shared Conceptual Ground
@@ -120,7 +129,7 @@ graph LR
     B -->|addresses| C
     A -->|employs| D
     B -->|contrasts with| D
-```
+````
 
 ```
 
@@ -152,3 +161,4 @@ graph LR
 
 #### Tags
 connect_ideas, vault_maintenance, knowledge_graph, cross_reference, analysis_tool
+```

.skills/skills/content-generation/content-generation/SKILL.md

@@ -12,6 +12,7 @@ This skill transforms academic or technical knowledge vault notes into polished,
 ## Target Platforms & Tone Adaptation
 
 **LinkedIn Posts** (150–300 words):
+
 - Conversational, first-person voice ("I've learned...", "Here's what I found...")
 - Hook opening: provocative question, surprising insight, or personal observation
 - 2–3 key takeaways, clearly structured
@@ -20,6 +21,7 @@ This skill transforms academic or technical knowledge vault notes into polished,
 - Tone: accessible, relatable, thought-leadership
 
 **Blog Posts** (600–1,500 words):
+
 - Narrative-driven introduction with real-world context
 - Structured sections with H2/H3 headers
 - Mix of explanation, examples, and actionable insights
@@ -29,6 +31,7 @@ This skill transforms academic or technical knowledge vault notes into polished,
 - Hashtags: not required; metadata/SEO keywords optional
 
 **Technical Articles** (800–2,000+ words):
+
 - Depth and nuance; cite sources and vault cross-references
 - Professional, authoritative tone (suitable for Medium, Dev.to, personal site)
 - Code snippets or architecture diagrams allowed
@@ -38,27 +41,32 @@ This skill transforms academic or technical knowledge vault notes into polished,
 ## Workflow
 
 ### 1. Input Detection
+
 - User selects text from a vault note or provides note reference
 - Agent asks (if unclear): "What format would you like? **LinkedIn post**, **blog article**, or **technical deep-dive**?"
 
 ### 2. Insight Extraction
+
 - Identify **2–3 most shareable, non-obvious insights** from the note
 - Prioritize: novel findings > practical lessons > conceptual clarity
 - Avoid generic statements; surface surprising or underexplored angles
 - Look for personal experience or lessons learned (Gianfranco's voice)
 
 ### 3. Structure & Adaptation
+
 - Rewrite content in target tone; remove academic formality
 - Create engaging opening hook (question, stat, or observation)
 - Restructure information for readability (short paragraphs, bullets for LinkedIn; narrative flow for blog)
 - Preserve technical accuracy; simplify jargon without losing precision
 
 ### 4. Call-to-Action & Engagement
+
 - LinkedIn: end with question, invitation to debate, or "What's your experience?"
 - Blog: conclusion with implications, link to related vault notes, or "What's next?"
 - Article: future directions, open problems, or invitation for feedback
 
 ### 5. Output & Polish
+
 - Generate markdown ready to copy-paste
 - Suggest hashtags for LinkedIn (include in output)
 - For blog/article: include title, optional subtitle, and metadata notes
@@ -67,18 +75,21 @@ This skill transforms academic or technical knowledge vault notes into polished,
 ## Example Workflow
 
 **Input**: Selection from `[[AI Safety/Prompt Injection Vulnerabilities]]` note
+
 ```
 [User selects technical explanation of prompt injection attack vectors]
 ```
 
 **Agent Detects**: Blog-post request (length and depth suggest longer form)
 
 **Agent Extracts**:
+
 - Insight 1: Most attacks succeed via **indirect injection** (document upload, search results), not direct user input
 - Insight 2: Defense requires **multi-layered approach** (parsing, sandboxing, monitoring), not single mitigation
 - Insight 3: Teams overlook **supply-chain injection** (third-party APIs as attack surface)
 
 **Agent Outputs** (Blog post structure):
+
 ```markdown
 # The Supply Chain Blind Spot in LLM Security
 
@@ -89,6 +100,7 @@ This skill transforms academic or technical knowledge vault notes into polished,
 ## The Indirect Injection Problem
 
 [Explanation of attack vectors beyond direct user input]
+
 - Document uploads
 - Web search results
 - Third-party API responses
@@ -119,6 +131,7 @@ This skill transforms academic or technical knowledge vault notes into polished,
 ## Output Format
 
 Return a markdown file with:
+
 1. **Title** and optional **Subtitle**
 2. **Metadata** (platform, word count, reading time, suggested hashtags)
 3. **Content** (ready to publish)