{"skill":{"slug":"intel-synthesis","displayName":"Intel Synthesis","summary":"Processes multi-source intelligence within 12-18h windows to cross-verify, deduplicate, and synthesize detailed geopolitical briefings with conflict analysis.","description":"# SKILL: intel-synthesis\n\n## Description\nAdvanced intelligence processing pipeline optimized for high-context models (Gemini 1.5 Pro/Ultra). Ingests raw multi-source data, performs cross-verification, deduplication, and conflict analysis, and produces authoritative geopolitical briefings.\n\n## Usage\nRun with the target directory path.\nExample: `openclaw run intel-synthesis --dir /Volumes/Intel/news/2026-02-13_0500`\n\n## Implementation\nThis skill leverages the large context window to process entire intel dumps in single passes, reducing hallucination and improving narrative coherence across multiple sources.\n\n### 1. Ingest & Analyze\n- Read MANIFEST.json to validate completeness.\n- Load all `.json` content files.\n- **Smart Temporal Filter:**\n  - Ali provides the files, but the *content* may contain old \"Related Articles\" or sidebars.\n  - **Rule:** Extract the *main narrative* event time.\n  - **Keep:** Events occurring within the assigned cycle (Last 12-18 hours).\n  - **Discard:** Obvious historical noise (e.g., \"2015\", \"10 years ago\") found in page clutter.\n  - *Note:* Do not auto-reject \"yesterday\" if it falls within the 12h operational window (e.g., late-night news for morning brief).\n\n### 2. Analyze (Single-Shot Chain of Thought)\n- **Persistence Check:** Read the most recent `[YYYY-MM-DD]_briefing_EN.md` from `/Volumes/Intel/NewsBriefs/`.\n- **Deduplication:** Identify stories that were already reported.\n- **Update Protocol:** \n  - If a story is new: Include as a full entry.\n  - If a story is an update: Frame it as an update (e.g., \"UPDATE: Following the PLC order reported in the previous cycle, UAE forces have begun withdrawing from...\") rather than repeating the baseline facts.\n  - If no new information exists for a previously reported story: Omit it to maintain high signal-to-noise.\n- **Entity Extraction**: Identify key actors, locations, and equipment.\n- **Timeline Reconstruction**: Order events chronologically, resolving \"2 hours ago\" vs absolute timestamps.\n- **Conflict Detection**: Highlight discrepancies between sources (e.g., casualty numbers, attribution).\n- **Source Weighing**: Prioritize primary/local sources over aggregators for facts; use aggregators for broad sentiment.\n\n### 3. Synthesize\n- **Core Objective:** Act as Chief Editor. Produce a cohesive narrative, not a summary. Prioritize strategic significance.\n- **Structural Instructions:**\n  - **Main Story (The Lead):** Select the single most significant geopolitical development. Format: Paragraphs as bullet points.\n  - **Theaters of Interest:** Lebanon, Syria, Palestine, Israel. Prioritize \"Key National Events\". Format: Bullet points (•).\n  - **Regional/International:** Group updates.\n- **Stylistic Guidelines:**\n  - **High-Density Writing:** Every sentence must carry weight.\n  - **Maximum Detail:** Extract maximum detail from source articles. Include specific names, dates, quantities, direct quotes, and operational context. Avoid generalization.\n  - **No Citations:** Do not include source tags (e.g., `[Reuters]`) in the text. The narrative must flow seamlessly as an authoritative voice.\n  - **Synthesize, Don't List:** Combine sources into fluid paragraphs.\n- **Temporal Filtering:** Strictly adhere to the 12-18 hour cycle window. Discard any articles or events that fall outside this timeframe, even if they appear in the source files.\n- **LaTeX Template:** All PDF generation must use the exact structure and commands from the official template located at `/Users/mikethebrain/.openclaw/workspace/sf.tex`. The agent must dynamically populate the content within this structure.\n- **Maximum Detail:** Extract maximum detail from source articles. Include specific names, dates, quantities, direct quotes, and operational context. Avoid generalization.\n- **No Citations:** Do not include source tags (e.g., `[Reuters]`) in the text.\n- **Synthesize, Don't List:** Combine sources into fluid paragraphs.\n\n### 4. Quality Assurance (Red Team)\n- **Self-Critique Cycle:** Before final output, review the draft for generic language (\"clashes\", \"tensions\").\n- **Resolution:** Replace vague terms with specific actions (e.g., \"artillery fire\", \"diplomatic protest\").\n- **Citation Check:** Verify every major claim has a source tag.\n\n### 5. Delivery\n- **PDF Generation:**\n  - Use `pandoc` or `wkhtmltopdf` if available? No, use **MacTeX** for professional typesetting.\n  - **IMPORTANT:** The `pdflatex` and `xelatex` binaries are located at `/Library/TeX/texbin/`. You must add this directory to your PATH or call them by full path.\n  - Command: `/Library/TeX/texbin/xelatex -output-directory=/Volumes/Intel/NewsBriefs/ [filename].tex`\n  - Ensure the `.tex` file uses `Arial Unicode MS` for Arabic support (`fontspec`) and sets direction to RTL.\n  - Pandoc command: `pandoc [input.md] -o [output.pdf] --pdf-engine=/Library/TeX/texbin/xelatex -V mainfont=\"Arial Unicode MS\" -V dir=rtl`\n- **Email Dispatch:**\n  - Construct email with `himalaya` using the configured account.\n  - Attach: `[Date]_briefing_EN.pdf`, `[Date]_briefing_AR.pdf`.\n  - Recipients: `abdayi@gmail.com`, `mdurankaddatz@gmail.com`, `corebrain2026@gmail.com`.\n  - Subject: `[Date] Daily Intelligence Briefing (Mike)`.\n\n## Configuration\n- **Model**: Requires a high-context model (Gemini 1.5 Pro, Claude 3 Opus/Sonnet) for best results.\n- **Safety**: Geopolitical content requires careful prompting to avoid safety refusals while maintaining neutrality.\n\n## Dependencies\n- `fs` for file reading.\n- `path` for handling directories.\n","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":343,"installsAllTime":13,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1771335860807,"updatedAt":1779077107006},"latestVersion":{"version":"1.0.0","createdAt":1771335860807,"changelog":"intel-synthesis 1.0.0\n\n- Initial release providing an advanced intelligence processing pipeline for high-context models.\n- Ingests and filters multi-source data, deduplicates stories, performs conflict and entity analysis, and reconstructs event timelines.\n- Synthesizes authoritative geopolitical briefings prioritized for strategic significance, following strict temporal windows and structural guidelines.\n- Generates professionally typeset PDFs using MacTeX with Arabic language and RTL support.\n- Automates delivery via email to preconfigured recipients with both English and Arabic briefings.","license":null},"metadata":null,"owner":{"handle":"mike-thebot","userId":"s17cr12fws8ar4bhwbf1gfhc1d885e6w","displayName":"Mike-thebot","image":"https://avatars.githubusercontent.com/u/259382515?v=4"},"moderation":{"isSuspicious":false,"isMalwareBlocked":false,"verdict":"clean","reasonCodes":["review.llm_review"],"summary":"Review: review.llm_review","engineVersion":"v2.4.24","updatedAt":1779972581774}}