Evaluate every source
CIA-style admiralty grading on every input. Source reliability and information credibility scored, defended, re-scored as new evidence arrives.
Training · Threat Intelligence · Strategic + Operational + Tactical
A practitioner CTI programme that trains the three levels of intelligence as distinct disciplines. Source evaluation. Diamond Model analysis. Attribution caution. Audience-fit reporting. Twelve weeks of structured analytic tradecraft, on real intel feeds and real engagement telemetry.
The CTI training gap
Real CTI is an analytic discipline. It evaluates sources, applies structured techniques, separates confidence from attribution, and writes for the right audience. Most curricula stop at "subscribe to the feed and import the STIX bundle." That is plumbing, not analysis.
01
Tactical IOCs, operational TTPs, and strategic posture get mixed in the same report. A board reads what an analyst should have read; an analyst gets language the board needed.
02
A free-tier feed and a partner-shared internal report get treated as equally trustworthy. Without admiralty-style grading, the analyst has no defensible confidence figure.
03
Public reporting routinely conflates "consistent with" with "attributed to". Practitioners absorb the bad habit. We drill the difference until it stops happening.
The three-level structure
The programme drills each level as a distinct discipline with its own consumer, its own tradecraft, its own product format — and the pipeline that connects them so a single evidence base feeds the analyst console, the operations centre, and the boardroom.
What you do on the programme
CIA-style admiralty grading on every input. Source reliability and information credibility scored, defended, re-scored as new evidence arrives.
ACH (Analysis of Competing Hypotheses). Key Assumptions Check. Devil's Advocacy. Red-team review. The toolkit professional analysts use to keep themselves honest.
TTPs at sub-technique grain. Diamond Model fills for adversary, infrastructure, capability, victim. Confidence figures defensible against peer review.
Estimative-language discipline (probable / likely / virtually certain). High-confidence does not mean "attributed". The two concepts live on different axes.
A boardroom briefing isn't a longer version of a SOC bulletin. Each register has its own length, vocabulary, structure, and what-not-to-say.
Every weekly intel product is workshopped in cohort review. Your sources, your confidence, your attribution claims — defended on the record.
Sample week · Source evaluation + analytic tradecraft · week 5 of 12
A representative tradecraft week. The work shifts from collecting evidence to grading it, weighing competing hypotheses against it, and defending your conclusions in cohort review.
Monday
Score every source on your desk against the NATO admiralty scale (A–F reliability, 1–6 credibility). Defend the grade you assigned the partner-shared report against the cohort.
Tuesday
Take a recent campaign reporting. Fill the Diamond Model — adversary, infrastructure, capability, victim — without conflating "consistent with" and "attributed to". Identify the corners that are weakly supported.
Wednesday
Build the ACH matrix for an ambiguous incident. List hypotheses, list evidence, score consistency. The point isn't to pick a winner — it's to make the evidence work harder.
Thursday
Take Wednesday's ACH. Surface every assumption it rested on. Run a red-team review with a different analyst. Watch your high-confidence conclusion become a medium-confidence one.
Friday morning
A short operational-level product on the incident from earlier in the week. Estimative language. Confidence stated. Attribution carefully phrased. Audience: SOC manager.
Friday afternoon
Every cohort member presents. Two peers play hostile reader. Your phrasing, your sourcing, your confidence figures are picked apart on the record. The misses are graded harder than the hits.
Tooling coverage
MISP, OpenCTI, and the TAW Threat Analyst Workbench. Ingestion, scoring, sharing, and pivot — read and write.
STIX 2.1 and TAXII 2.1 at byte level. Author bundles. Read other people's bundles. Argue about their indicator-pattern syntax.
Maltego transforms. SpiderFoot. Shodan. Censys. Public dataset triage with command-line tools.
CERT-In / NCIIPC advisory, Diamond Model and MITRE ATT&CK + Navigator. Used to structure analysis, not as marketing labels.
ACH. Key Assumptions Check. Devil's Advocacy. Pre-mortem. Outside View. Memorised, drilled, applied weekly.
CERT-In TLP, NCIIPC CII categorisation. NATO Admiralty grading and ICD 203 estimative language as the international tradecraft baseline. Each used until it is second nature.
A structured intel-product template library covering tactical bulletins, operational briefs, and strategic memos. Audience-specific style guides.
Citation manager workflow. Source-corroboration matrices. The boring craft that makes an intel product survive a regulator review.
What you walk away with
A tactical bulletin, an operational brief, a strategic memo, and weekly analytic products. Show them at interview.
You can show the source grades, the ACH matrices, and the key-assumption checks behind every conclusion you reach.
You enter your first analyst role already fluent in the workbench paradigm (see Products → TAW). One less three-month ramp.
Direct routes into BFSI threat-intel teams, sector-CSIRT analyst roles, sovereign intel cells, MSSP CTI desks.
Drilled against
Practice
"Confidence is a number. Attribution is an argument. The day you stop confusing the two is the day you start writing intel products that survive a regulator review."
Weekend and weekday cohorts. Corporate cohorts on request. The weekly intel-product workload is real; budget eight to ten hours per week outside the cohort sessions.