Quick Ref / Command

Engineer Kit

/ak:cti-expert

Cyber threat intelligence and OSINT

Cyber threat intelligence and OSINT analysis for exposure reviews, recon, breach checks, forensics, cloud tenant recon, threat modeling, and structured reports.

01

Acquire

02

Enrich

03

Assess

04

Deliver

Rule 01

Use public data only and preserve legal/ethical boundaries.

Rule 02

Every claim needs a source, trust score, and confidence level.

Rule 03

Tool limitations are collection gaps, not case blockers.

Rule 04

ASCII visualizations are the default for portable reports.

Rule 05

Deliver Markdown and DOCX for reports and briefs unless the format is machine-only.

Execution Map

Core lanes from input to output, aligned with the command's real execution path.

01

Start

Input, scope, route

  1. 1 Acquire Collect public raw data for the bounded target with /case, /sweep, /query, username, phone, email-deep, subdomain, breach, traffic, techstack, secrets, threat-check, and target-specific techniques as applicable.
02

Work

Agent / skill execution

  1. 2 Enrich Expand discovered identifiers with /branch, /crossref, /link-subjects, /timeline, /show-trail, signatures, and parallel AgentFlow enrichment when three or more subjects exist.
03

Verify

Gate, review, validation

  1. 3 Assess Score exposure and threat model findings, then run /validate, /coverage, /verify-finding, blind-spot checks, and conflict handling so contested evidence remains visible.
04

Close

Report, handoff, artifact

  1. 4 Deliver Package cited intelligence with /report, /brief, /render, or /workspace save; /case, /report, and /brief auto-save Markdown plus DOCX, with optional HTML mirror when requested.

Syntax, arguments, subcommands

How to invoke the skill: syntax, positional arguments, shared options, then each subcommand with its own syntax and outcome.

Syntax

/ak:cti-expert [target] [--yolo] [--sequential] [--format html] [--no-antv|--no-diagram-design|--no-editorial-visuals]

Arguments

[target]

Investigation target

Permitted public identifier such as a domain, IP, email, username, phone number, image, organization, CVE, or threat indicator. Keep authority, purpose, allowed sources, retention, and contact boundaries explicit.

/ak:cti-expert "Run a case on own-domain.example. Use public passive sources only, label inference, redact personal data, install nothing, and stop before contacting any account or system."

Shared options

--yolo

Skip prompts

Runs without clarification or confirmation prompts. It does not approve new scope, paid providers, credential use, active probing, dependency installation, or publication.

--sequential

Disable parallel enrichment

Disables packaged parallel enrichment. Case operations such as /case, /sweep, /query, and /flow run inside the CTI workspace; they are not AgentKit skill flags.

--format html

HTML mirror

Adds an optional self-contained HTML mirror beside the default Markdown and DOCX auto-save. Docs do not document --format md as a skill flag.

Subcommands

case Full case
/case [target]

Run the full case pipeline inside the CTI workspace for the supplied target.

[target]
Public identifier to investigate within the approved boundary.

Outcome: Cited findings, subject graph, confidence and coverage notes, and saved Markdown plus DOCX reports when conversion tools are available.

/case example.com
sweep Multi-vector sweep
/sweep [target]

Collect broad public records and observations for any target type.

[target]
Identifier to collect leads for before choosing pivots.

Outcome: Raw leads with collection method, source, confidence, and collection gaps recorded.

/sweep @username
query Operator queries
/query [subject]

Generate advanced search-operator queries for public-source collection.

[subject]
Domain, name, handle, organization, or other subject to query.

Outcome: A query set ready for acquire-stage collection, with uncertainty and source limits preserved.

/query example.com
flow Guided flow
/flow [type]

Start the first-time guided workflow for a target category.

[type]
Workflow such as person, domain, email, or quick.

Outcome: Step-by-step prompts that keep purpose, sources, confidence, and safety checks explicit.

/flow person
progress Progress
/progress

Show the current case phase and what remains pending.

Outcome: Current acquire, enrich, assess, or deliver status with pending work.

/progress
validate Validate findings
/validate

Audit case quality and evidence completeness; it is a workflow check, not proof that every claim is true.

Outcome: Quality score and issues to revisit before final delivery.

/validate
coverage Coverage matrix
/coverage

List sources, target areas, and blind spots that were or were not covered.

Outcome: Coverage matrix with collection gaps recorded as limitations.

/coverage
brief Brief
/brief

Produce a plain-language summary for non-technical readers.

Outcome: Concise cited brief, saved as Markdown and DOCX when conversion tools are available.

/brief
report Report export
/report [brief|json|csv|legal|journalist|ioc]

Export the case in the selected report format.

[brief|json|csv|legal|journalist|ioc]
Optional output style. Omit for the formal structured intelligence report.

Outcome: Selected report artifact with citations, confidence, contradictions, and unresolved gaps. Markdown and DOCX are saved for narrative reports; machine exports are separate.

/report legal
render Render view
/render <entities|timeline|risk|network|threat-path|attack-surface>

Render a portable ASCII relationship, timeline, risk, network, threat-path, or attack-surface view unless Mermaid is explicitly requested.

<entities|timeline|risk|network|threat-path|attack-surface>
Visualization to render from current case data.

Outcome: ASCII visualization suitable for Markdown and DOCX reports.

/render entities
workspace Workspace
/workspace <save|open|list|diff> [name]

Save, resume, list, or compare CTI case workspaces.

<save|open|list|diff>
Workspace operation to run.
[name]
Case workspace name or pair of names for diff.

Outcome: Case state is saved, resumed, listed, or compared without changing the evidence requirements.

/workspace save mycase

Output / publish flags

Only flags that change the artifact or publish it. Not the full option set.

--no-antv

Disable AntV

Disable preferred AntV chart rendering for HTML mirrors.

Example

/ak:cti-expert own-domain.example --format html --no-antv
--no-diagram-design

Disable diagram-design

Disable diagram-design layouts for HTML diagrams.

Example

/ak:cti-expert own-domain.example --format html --no-diagram-design
--no-editorial-visuals

Disable editorial visuals

Disable editorial visual enhancements for HTML output.

Example

/ak:cti-expert own-domain.example --format html --no-editorial-visuals

Sample Prompt

Concrete invocations for each flag, subcommand, mode, or route available in the live workflow.

Bounded case Recommended
/ak:cti-expert "Run a case on own-domain.example. Use public passive sources only."
Use when:
Run a permitted OSINT or threat-intelligence investigation on a bounded target.
Expected:
Starts the CTI workspace. Full-case, sweep, query, and guided flow are in-run operations such as /case, not /ak:cti-expert --case flags.
Sequential enrichment
/ak:cti-expert own-domain.example --sequential
Use when:
Disable packaged parallel enrichment for a permitted public target.
Expected:
Runs without packaged parallel enrichment while preserving ethics, citations, coverage checks, and evidence confidence.
HTML mirror
/ak:cti-expert own-domain.example --format html
Use when:
Keep default Markdown and DOCX output and add an HTML mirror.
Expected:
Writes the default report package plus an optional self-contained HTML mirror with cited intelligence.
Skip prompts
/ak:cti-expert own-domain.example --yolo
Use when:
Skip clarification prompts without expanding authority.
Expected:
Runs without extra prompts. It does not approve paid providers, credentials, probing, or publication.
In-run sweep
/sweep @username
Use when:
Use inside a CTI workspace to collect broad public leads before pivoting.
Expected:
Records raw leads with source, collection method, confidence, uncertainty, and visible collection gaps.
Report export
/report legal
Use when:
Use inside a CTI workspace when findings need a legal evidence format.
Expected:
Exports cited findings, confidence, contradictions, and unresolved gaps in the selected legal report artifact.

Handled Scope

  • Domain and subdomain recon
  • Breach and leak checks
  • Username/email/phone OSINT
  • Image and document forensics
  • M365/Azure tenant recon
  • Exposure scoring
  • Threat modeling
  • Structured intelligence reporting

Artifacts Produced

CTI report package

Pattern

OSINT-REPORT-[CASE-ID]-[YYYY-MM-DD].md + .docx; optional CTI-REPORT-[CASE-ID]-[YYYY-MM-DD].html mirror

Reports preserve narrative evidence, structured JSON for charts, source citations, confidence, gaps, and recommendations.

Related skills

Next