Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add AppliedIR/sift-mcp --skill vhir-welcomegit clone --depth 1 https://github.com/AppliedIR/sift-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/skills/appliedir/sift-mcp/vhir-welcome)<a href="https://agentmods.dev/skills/appliedir/sift-mcp/vhir-welcome"><img src="https://agentmods.dev/badge/skills/appliedir/sift-mcp/vhir-welcome/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/skills/appliedir/sift-mcp/vhir-welcome"><img src="https://agentmods.dev/badge/skills/appliedir/sift-mcp/vhir-welcome.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 | $0.00059 | $0.02001 |
| Opus 5.5 | $0.00024 | $0.00800 |
| Sonnet 5.5 | $0.00012 | $0.00400 |
| Haiku 4.5 | $0.00006 | $0.00200 |
Grade B, and why
vhir-welcome scanned grade B with 1 finding 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 4d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
Read settings.json (both `~/.claude/settings.json` and project How it starts
The opening of the file, as written. The whole thing — 240 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Post-Install Verification and Onboarding
Follow these phases in order. Report results as you go. Stop on failure.
Phase 1: Your Arsenal
Step 1: Backend Inventory
Run vhir setup test via Bash. If vhir is not in PATH (common on
fresh installs before shell restart), use ~/.vhir/venv/bin/vhir
instead. Do NOT warn about PATH — the installer already told the user
to restart their shell.
It hits the gateway /health endpoint and returns per-backend status
with tool counts. Report the results.
If the gateway is not responding, stop immediately:
"Gateway not reachable. Check: vhir service status"
Step 2: Forensic Capabilities
Call list_available_tools on sift-mcp. Do NOT display the raw tool list.
Summarize as a single line with count and categories:
Forensic tools: N available (memory, filesystem, registry, timeline, network, imaging)
Only mention tools that are MISSING from expected categories. Do not list individual tools unless they are absent. Do NOT show install commands for missing tools — missing tools are a SIFT distribution issue, not a Valhuntir problem.
Zimmerman tools that only run on Windows (PECmd, SrumECmd) are not expected on Linux. Do not flag them as missing. If the examiner needs prefetch or SRUM parsing, mention WinTools MCP as the solution.
Step 3: Knowledge and Baselines
Call get_stats on forensic-rag and get_health on windows-triage:
Knowledge base: N records across M sources
Triage baselines: known_good.db (X entries), context.db (Y entries)
Step 4: Report Profiles
Call list_profiles on report-mcp. List profiles with descriptions
as returned by the tool.
Phase 2: How to Work
Step 5: The Interaction Model
Explain:
Valhuntir is conversation driven. Ask Claude to do anything:
- "Create a case called incident-001"
- "Register this evidence file"
- "Run volatility on this memory dump"
- "What artifacts indicate credential dumping?"
- "Generate an executive report"
Claude handles almost everything through MCP tools. No need to
memorize lots of commands or tool names.
The only CLI-only operations:
vhir approve Approve findings (signed with your key; your passphrase unlocks it)
vhir reject Reject findings with reason (signed the same way)
vhir exec Run forensic command with TTY confirmation
vhir config --setup-password Set your passphrase and signing key
vhir config --reset-password Change your passphrase
These are for YOU at the terminal. Claude's settings deny it vhir approve
and vhir reject, and your passphrase is read from the terminal.
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.
- 4d ago First seen · 240 lines · 59 tokens per session scan B d2d20f1d09a2
vhir-welcome is a skill published in the GitHub repository AppliedIR/sift-mcp (18 stars, last pushed today), licensed MIT. It adds 59 tokens to every session and 2,001 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-12.
Other skills, from other repositories
performing-cloud-forensics-with-aws-cloudtrail
Perform forensic investigation of AWS environments using CloudTrail logs to reconstruct attacker activity, identify compromised credentials, and analyze API call patterns.
collecting-volatile-evidence-from-compromised-host
Collect volatile forensic evidence from a compromised system following order of volatility, preserving memory, network connections, processes, and system state before they are lost.
agent-artifact-catalog
Research, author, and validate AI agent artifact catalog entries documenting the forensic artifacts AI agents leave on endpoints - install paths, config and credential files, MCP server configs, listening ports, process trees, registry keys, and the Windows event log records that prove a tool ran. Use this skill…
collecting-volatile-evidence-from-a-suspect-host
Collects volatile evidence from a potentially compromised host in correct order of volatility: memory, network state, running processes, logged-on users, and open handles, while preserving integrity and minimizing footprint. Activates for requests to collect volatile evidence, perform live response, or capture host…
windows-event-logs-analysis
Conduct comprehensive forensic analysis of Windows Event Logs (.evtx) to trace attacker activity including lateral movement, privilege escalation, credential dumping, persistence mechanisms, and remote code execution. Use this skill during incident response, threat hunting, or post-breach forensic investigations on…
windows-prefetch-analysis
Analyze Windows prefetch files (.pf) to determine evidence of program execution. This skill details how to extract execution times, run counts, and the paths of files accessed by a program, which is critical for incident response and malware timeline reconstruction.