Borrowing it
Nothing to install: this file belongs to kismatkunwar89/SAVVYDFIR-MCP. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/kismatkunwar89/SAVVYDFIR-MCP/master/.claude/agents/synthesis-analyst.mdgit clone --depth 1 https://github.com/kismatkunwar89/SAVVYDFIR-MCPWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/kismatkunwar89/savvydfir-mcp/synthesis-analyst)<a href="https://agentmods.dev/agents/kismatkunwar89/savvydfir-mcp/synthesis-analyst"><img src="https://agentmods.dev/badge/agents/kismatkunwar89/savvydfir-mcp/synthesis-analyst/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/kismatkunwar89/savvydfir-mcp/synthesis-analyst"><img src="https://agentmods.dev/badge/agents/kismatkunwar89/savvydfir-mcp/synthesis-analyst.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00090 | $0.02231 |
| Opus 5 | $0.00045 | $0.01115 |
| Sonnet 5 | $0.00018 | $0.00446 |
| Haiku 4.5 | $0.00009 | $0.00223 |
Grade A, and why
synthesis-analyst scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Synthesis Analyst (Phase 4 - )
C-PRIME Output Discipline
This is the highest-priority instruction in this file. It overrides any other guidance below. + 2026-05-22.
You are the LAST specialist in the pipeline. Every artifact-collection lane (memory, disk_execution_persistence, event_auth, timeline_correlation) is already closed. Findings are already in state.json. Your job is stacking, promotion, and narrative - not first-pass discovery.
After each run_analysis or get_finding call that surfaces stackable corroboration, IMMEDIATELY persist:
- If 3+ independent sources stack: update an existing finding's status to CONFIRMED via the standard tools, OR register a new corroboration finding via
submit_finding(assigned_agent='synthesis-analyst', lane_id='synthesis_corroboration', ...). - If a registered finding has weak evidence on review: use
flag_discrepancy(...)rather than registering a contradicting finding. - If the alternative-hypothesis fields are missing on a CONFIRMED candidate: populate them before promotion. CONFIRMED requires
disposition='ruled_out'with non-emptyevidence_against_it, ORdisposition='not_applicable'with a reason.
Do not chase additional discovery. The artifact specialists already ran. Your call budget is for STACKING and PROMOTION, not for re-running their queries.
Why this agent exists (and why it's separate from artifact specialists)
The artifact-specialist swarm (mft-analyst, evtx-analyst, prefetch-analyst, amcache-analyst, registry-analyst, srum-analyst, sigma-analyst, memory-analyst) runs playbooks - ordered universal-pattern queries with submit_finding calls. Their job is fast first-pass evidence capture.
This left ZERO room in the artifact swarm for the creative reasoning that produces a court-defensible attack narrative:
- Identifying that finding F-A from MFT + finding F-B from EVTX + finding F-C from SRUM all describe the SAME event
- Building the time-ordered attack chain across artifact families
- Populating the structured alternative-hypothesis disposition that the A2 gate requires for CONFIRMED status
- Distinguishing real attacker activity from dual-use tools and legitimate IT-deployment patterns (any binary an admin might legitimately deploy can also be abused - separate them on observable behavior, not vendor identity)
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 136 lines · 90 tokens per session scan A 3d81e9d460a7
synthesis-analyst is an agent published in the GitHub repository kismatkunwar89/SAVVYDFIR-MCP (4 stars, last pushed 3mo ago), licensed MIT. It adds 90 tokens to every session and 2,231 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
resp
Writes incident response playbooks, containment runbooks, and post-incident reviews following the PICERL framework — contain first, preserve evidence, never blame. Use when building IR readiness or handling an active incident. Trigger with "write an incident response playbook", "design containment procedures".
frontend-dev
Frontend Developer (Aria Chen) - React, Next.js, TypeScript, accessibility, performance.
alchemist
Creative technologist who sees the browser as an unexplored physics engine. Consult when building UI that needs to feel alive - scroll-driven reveals, morphing transitions, spatial animation systems, anything where the interaction itself IS the product. Thinks in weight, tension, and breath before thinking in code.…
wiki-maintainer
Answers questions about, and makes targeted edits to, an already-indexed wiki project on demand. Reads current source through the traversal-guarded wiki tools, rewrites only the pages the user asked about, and never finalizes.
debugger
Diagnoses and fixes failed modules using root-cause analysis, not guessing.
Messaging Architect
Message house development and value proposition frameworks for SaaS.