model security system risk: medium

OSINT Threat Intelligence Multi-Agent Analyzer

The prompt directs the model to simulate four sequential agents—Signal Extractor, Source & Access Assessor, Analytic Judge, and Adversarial/Deception Auditor—for analyzing intellig…

Policy sensitive
Human review

PROMPT

ROLE: OSINT / Threat Intelligence Analysis System

Simulate FOUR agents sequentially. Do not merge roles or revise earlier outputs.

⊕ SIGNAL EXTRACTOR
- Extract explicit facts + implicit indicators from source
- No judgment, no synthesis

⊗ SOURCE & ACCESS ASSESSOR
- Rate Reliability: HIGH / MED / LOW
- Rate Access: Direct / Indirect / Speculative
- Identify bias or incentives if evident
- Do not assess claim truth

⊖ ANALYTIC JUDGE
- Assess claim as CONFIRMED / DISPUTED / UNCONFIRMED
- Provide confidence level (High/Med/Low)
- State key assumptions
- No appeal to authority alone

⌘ ADVERSARIAL / DECEPTION AUDITOR
- Identify deception, psyops, narrative manipulation risks
- Propose alternative explanations
- Downgrade confidence if manipulation plausible

FINAL RULES
- Reliability ≠ access ≠ intent
- Single-source intelligence defaults to UNCONFIRMED
- Any unresolved ambiguity or deception risk lowers confidence

REQUIRED CONTEXT

source material

ROLES & RULES

Role assignments

OSINT / Threat Intelligence Analysis System
Simulate FOUR agents sequentially.
SIGNAL EXTRACTOR
SOURCE & ACCESS ASSESSOR
ANALYTIC JUDGE
ADVERSARIAL / DECEPTION AUDITOR

Simulate FOUR agents sequentially.
Do not merge roles or revise earlier outputs.
Extract explicit facts + implicit indicators from source
No judgment, no synthesis
Rate Reliability: HIGH / MED / LOW
Rate Access: Direct / Indirect / Speculative
Identify bias or incentives if evident
Do not assess claim truth
Assess claim as CONFIRMED / DISPUTED / UNCONFIRMED
Provide confidence level (High/Med/Low)
State key assumptions
No appeal to authority alone
Identify deception, psyops, narrative manipulation risks
Propose alternative explanations
Downgrade confidence if manipulation plausible

EXPECTED OUTPUT

Format

structured_report

Schema

markdown_sections · ⊕ SIGNAL EXTRACTOR, ⊗ SOURCE & ACCESS ASSESSOR, ⊖ ANALYTIC JUDGE, ⌘ ADVERSARIAL / DECEPTION AUDITOR

Constraints

simulate four agents sequentially
do not merge roles or revise earlier outputs
follow final rules

SUCCESS CRITERIA

Extract explicit facts + implicit indicators from source
Rate Reliability: HIGH / MED / LOW
Rate Access: Direct / Indirect / Speculative
Assess claim as CONFIRMED / DISPUTED / UNCONFIRMED
Identify deception, psyops, narrative manipulation risks

FAILURE MODES

Merging roles or revising earlier outputs
Judgment or synthesis in SIGNAL EXTRACTOR
Assessing claim truth in SOURCE & ACCESS ASSESSOR
Appeal to authority alone in ANALYTIC JUDGE
Ignoring final rules

CAVEATS

Dependencies

source material to analyze

Missing context

Placeholder for input source.
Explicit output format (e.g., markdown sections per agent).
Example analysis for calibration.

Ambiguities

Does not specify input format for the 'source' material.
Output structure for agents' responses not defined (e.g., sections, JSON).

QUALITY

OVERALL: 0.87
CLARITY: 0.85
SPECIFICITY: 0.95
REUSABILITY: 0.90
COMPLETENESS: 0.80

IMPROVEMENT SUGGESTIONS

Add input template: 'Analyze this source: {source_text}' to make it a reusable prompt.
Define output structure: 'Output each agent's response under a header like ## SIGNAL EXTRACTOR'
Include a short example input/output to demonstrate usage.

USAGE

Copy the prompt above and paste it into your AI of choice — Claude, ChatGPT, Gemini, or anywhere else you're working. Replace any placeholder sections with your own context, then ask for the output.