Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/depalmar/ai-dfir-toolkitnpx agentmods add skills/depalmar/ai-dfir-toolkit/agent-artifact-catalogWrote 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/depalmar/ai-dfir-toolkit/agent-artifact-catalog)<a href="https://agentmods.dev/skills/depalmar/ai-dfir-toolkit/agent-artifact-catalog"><img src="https://agentmods.dev/badge/skills/depalmar/ai-dfir-toolkit/agent-artifact-catalog/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/depalmar/ai-dfir-toolkit/agent-artifact-catalog"><img src="https://agentmods.dev/badge/skills/depalmar/ai-dfir-toolkit/agent-artifact-catalog.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00197 | $0.03589 |
| Opus 5.5 | $0.00079 | $0.01436 |
| Sonnet 5.5 | $0.00039 | $0.00718 |
| Haiku 4.5 | $0.00020 | $0.00359 |
Grade A, and why
agent-artifact-catalog 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 24d 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 — 316 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI Agent Artifact Catalog Authoring
This catalog documents what AI agents leave behind on an endpoint so responders can find, collect, and interpret it. That framing matters: it is a collection and analysis reference, not an abuse-technique reference. LOL-family projects answer "what can this be misused to do." This one answers "what is on the disk, what does it prove, and in what order do I collect it."
Keep that distinction in view while authoring. If a proposed field describes an
attack technique rather than an observable artifact, it belongs in the
abuse_potential prose field or in the ATLAS mapping, not as a new artifact type.
Layout
The catalog lives at artifacts/ inside the toolkit. Run every command from
that directory.
artifacts/
catalog/ one YAML file per tool, slug-named
case-studies/ documented real-world incidents with IOC blocks
schema/ artifact.schema.json + entry-template.yml
scripts/ validate.py, export.py
docs/api/ generated JSON/CSV feeds - never hand-edit
docs/ VERIFICATION.md, MCP.md
Authoring workflow
Work in this order. Each step feeds the next, and skipping research produces entries that look authoritative but are guesses.
1. Check for an existing entry
grep -ril "<tool name>" catalog/
Update in place rather than adding a duplicate. IDs are permanent once merged - never renumber an existing entry, because downstream detections reference them.
2. Research the artifacts
Prefer sources in this order, because confidence ratings depend on provenance:
- A real installation you can inspect. Install in a VM, run it, then diff the
filesystem and enumerate listeners. This is the only path to
confidence: highon a path you have not seen documented. - Vendor documentation - official docs, the source repo, release notes.
- Source-code reading - config loaders and path constants in the repo.
- Third-party analysis - sandbox reports, security research.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 24d ago Changed · +23 lines aeff5c5ee0d3
- 1mo ago First seen · 293 lines · 197 tokens per session scan A bc245301afb5
agent-artifact-catalog is a skill published in the GitHub repository depalmar/ai-dfir-toolkit (129 stars, last pushed 6d ago), licensed Apache-2.0. It adds 197 tokens to every session and 3,589 once invoked, about $0.0008 per session on Opus 5.5. 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 2026-08-30.
Other skills, from other repositories
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…
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.
vhir-tool-reference
Which Valhuntir MCP tool or SIFT tool to use for a question, file, IOC or artifact, and the step-by-step workflows that combine them (IOC investigation, threat actor research, detection engineering, malware analysis, disk, memory, timeline and network). Use when choosing tools or planning a multi-step analysis.
finding-skeptic
Use before calling recordfinding on forensic-mcp, to challenge a drafted forensic finding. An isolated skeptic re-runs the cited tools and tries to refute the observation and interpretation; the investigator then records the finding as upheld, revised, contested or withdrawn.
vhir-ingest
How to start an investigation once the evidence is surveyed: which tools to run first, ingesting evidence into OpenSearch (triage packages, memory, delimited files) and querying it. Use after the evidence survey, before analysis, or when ingesting or querying indexed evidence.