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/vectra-ai-research/vectra-soc-agent-starternpx agentmods add skills/vectra-ai-research/vectra-soc-agent-starter/vectra-reportsWrote 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/vectra-ai-research/vectra-soc-agent-starter/vectra-reports)<a href="https://agentmods.dev/skills/vectra-ai-research/vectra-soc-agent-starter/vectra-reports"><img src="https://agentmods.dev/badge/skills/vectra-ai-research/vectra-soc-agent-starter/vectra-reports/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/vectra-ai-research/vectra-soc-agent-starter/vectra-reports"><img src="https://agentmods.dev/badge/skills/vectra-ai-research/vectra-soc-agent-starter/vectra-reports.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 | $0.00195 | $0.02989 |
| Opus 5 | $0.00097 | $0.01494 |
| Sonnet 5 | $0.00039 | $0.00598 |
| Haiku 4.5 | $0.00019 | $0.00299 |
Grade A, and why
vectra-reports 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 12d 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 — 227 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Vectra AI Reports (Python channel)
Render canned, named, repeatable dashboards against a Vectra AI tenant.
Each report is a YAML definition under definitions/ that bundles one or more
SQL queries against the Investigation Query API plus a rendering spec (summary
KPIs, tables, pie charts, Sankey diagrams). Queries inside a single report run
in parallel, gated only by the documented 5 req/min token bucket.
When to use this skill
Use only when the user explicitly names a canned report from the
catalog (run python scripts/list_reports.py to see it). Examples:
- "Run the C2 beacon report for the last 24 h"
- "Render the top-talkers dashboard as HTML"
- "Show me the DNS error rate report, last hour"
- "Generate the zone-to-zone data transfer report"
- "Give me the TLS posture report as Markdown"
The trigger is the report name, not the data domain. A report
can be named by its exact ID (protocol_distribution) or by its
catalog label phrased as a question ("what's the protocol
distribution across the network?", "show me active connections
right now"). Both count as a named trigger.
When NOT to use this skill
Reports are dashboards, not investigation tools. If the request is
investigative ("check CloudTrail", "what did this account do", "who's
behind this IP", "pivot from detection <id>", "investigate entity
<name>", "find Kerberoasting last 7 d", "sweep this CISA advisory"),
route to vectra-hunt instead.
Full routing table lives in
reference/ROUTING.md — shared with
vectra-reports-mcp, so update it there.
If the user names neither a specific report nor a clear investigation
question, list the available reports (python scripts/list_reports.py)
and ask them to pick one — do not silently default to a generic report.
Channel selection — Python vs MCP
This skill (vectra-reports) is the Python channel. The same report
catalogue is also runnable via the MCP channel
(vectra-reports-mcp) which needs no
Python venv. Pick one channel per task and stick with it; do not mix
mid-run. If the Python venv (3.11+) isn't available, switch to the MCP
channel — do not hand-roll REST calls against the Investigation
Query API. The MCP server handles auth, polling, rate limits, and
response shape; bypassing it is a known source of cascading failures
(OAuth body vs Basic, polling endpoint shape, request-id lifecycle).
See reference/ROUTING.md for the channel
selection rules.
What ships with it
40 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.
- .env.example 853 B
- definitions/active_connections.yaml 3.1 KB
- definitions/c2_beacon_report.yaml 1.9 KB
- definitions/cert_expiration.yaml 3.0 KB
- definitions/daily_threat_summary.yaml 2.9 KB
- definitions/dns_error_rate.yaml 2.7 KB
- definitions/flow_records.yaml 4.7 KB
- definitions/http_status_codes.yaml 3.1 KB
- definitions/npm_active_tcp_udp_connections.yaml 2.0 KB
- definitions/protocol_distribution.yaml 3.2 KB
- definitions/remote_access_sessions.yaml 6.2 KB
- definitions/rpc_latency.yaml 3.4 KB
- definitions/saas_reachability.yaml 5.4 KB
- definitions/tls_session_duration.yaml 3.6 KB
- definitions/top_listeners_by_ip.yaml 2.6 KB
- definitions/top_talkers_senders_by_ip.yaml 2.2 KB
- definitions/vlan_utilization.yaml 3.6 KB
- definitions/zone_segment_data_transfers.yaml 3.0 KB
- engine/__init__.py 1.1 KB runs code
- engine/client.py 13 KB runs code
- engine/executor.py 11 KB runs code
- engine/formatters.py 2.9 KB runs code
- engine/loader.py 7.0 KB runs code
- engine/renderer.py 23 KB runs code
- engine/templates/base.html 1.8 KB
- engine/templates/pie_chart.html 1.9 KB
- engine/templates/summary.html 1.5 KB
- engine/templates/table.html 2.4 KB
- INSTALL.md 3.0 KB
- pyproject.toml 608 B
- reference/AUTHORING.md 10 KB
- reference/ROUTING.md 3.2 KB
- reference/SQL_GUIDE.md 4.7 KB
- scripts/_path.py 419 B runs code
- scripts/list_reports.py 2.2 KB runs code
- scripts/run_report.py 3.7 KB runs code
- scripts/show_report.py 2.9 KB runs code
- scripts/validate.py 1.6 KB runs code
- tests/test_executor.py 2.2 KB runs code
- uv.lock 82 KB
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.
- 12d ago First seen · 227 lines · 195 tokens per session scan A 45c884584e0d
vectra-reports is a skill published in the GitHub repository vectra-ai-research/vectra-soc-agent-starter (2 stars, last pushed 3d ago), licensed MIT. It adds 195 tokens to every session and 2,989 once invoked, about $0.0010 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 skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…