Seven-Question Assessment Reference
These questions assess different dimensions of a finding. Technical validity depends on independent reproduction and evidence; scope, program eligibility, duplicate status, and presentation quality are recorded separately. The active workflow owns sequencing, report completion, and submission disposition. A negative answer is not by itself a technical rejection.
The 7 questions
1. Is the asset in scope?
Check the engagement's scope.md (recon/decepticon's RoE doc):
- In-scope domain list
- In-scope IP/CIDR list
- Out-of-scope explicit exclusions (test envs, staging, third-party CDNs/CDN-managed subdomains where the org doesn't own the underlying machine)
An asset outside approved scope must not receive target-facing testing. Record the scope constraint and policy source separately from any existing technical evidence; scope exclusion does not refute that evidence.
2. Is there real-world impact?
Distinguish demonstrated harm from hypothetical consequences. Concrete impact statements include:
- "An attacker can read victim user X's PII (email, name, DOB, SSN)"
- "An attacker can post on behalf of victim user X"
- "An attacker can transfer funds from victim user X's account"
- "An attacker can persist code on the production server"
- "An attacker can pivot to internal network 10.0.0.0/8"
If the finding's impact is "configuration is non-default" or "the manual recommends X but the deployment does Y" without concrete attacker-reachable harm, record the observed behavior and impact limitation without claiming an unsupported vulnerability or severity.
3. Does the PoC actually prove the impact?
The validate_finding result is necessary but not sufficient. The
PoC must demonstrate the IMPACT, not just trigger the vector.
Examples:
- IDOR PoC must show ATTACKER session reading VICTIM data — not just "request returned 200 OK"
- XSS PoC must execute attacker-controlled JS in victim's browser context — not just "alert(1) reflected in HTML source"
- SQLi PoC must extract real data — not just "single quote → 500"
- SSRF PoC must reach an internal-only target — not just "external fetch worked"
If the PoC stops short of the claimed impact, that impact remains unverified. Describe the missing proof and preserve any narrower demonstrated result. A false-positive verdict requires evidence refuting the technical claim.
4. Is the impact above the program's severity floor?
Many BB programs explicitly out-of-scope:
- CSRF on logout endpoint
- Self-XSS (requires victim to inject own payload)
- Missing security headers
- Information disclosure of public-by-design info
- Rate-limit issues w/o demonstrated abuse
- Subdomain takeover candidates where ownership can't be proven
- Clickjacking without authenticated state change
Compare against the supplied program's "out of scope" / "won't fix" / "informational only" list and cite the policy source. A finding below the floor may remain technically verified. Record eligibility separately; the floor does not justify changing CVSS metrics or expanding testing to raise severity.
5. Can the operator reproduce it from your PoC alone?
A triager will not have your engagement context. The PoC must work standalone:
- All required prerequisites stated explicitly (account registration, specific user role, specific test data)
- Exact URL / parameter / cookie values
- Browser version if it matters
- Date/time stamp if the bug is recent and the program may patch in between
If the PoC requires "you also need state X that I had set up", document that setup and the missing reproduction evidence. Incomplete instructions are a report-quality defect; insufficient independent reproduction leaves the claim unverified rather than automatically refuted.
6. Is it a known duplicate?
Compare supplied prior reports, CHANGELOG entries, recent disclosures, and
local confirmed-findings.md / rejected-hypotheses.md records by mechanism,
component/version, privileges, and supported impact. Cite the matching evidence;
missing history means unknown novelty.
A duplicate remains a technical finding. Record exact overlap or a materially different variant separately from verification status. This assessment does not authorize external account access or submission.
7. Does the title state the supported impact in one line?
Bad titles:
- "SSRF on /webhook"
- "JWT issue"
- "Mass assignment"
Good titles:
- "SSRF on /webhook → AWS instance-metadata cred extract → full account takeover"
- "JWT alg=none bypass → admin impersonation of any user"
- "Mass assignment on PATCH /api/users/me → self-promote to is_admin=true"
A useful title names {vuln class, target, impact, severity} to the extent supported by evidence. Revise a weak title without changing the technical verdict; do not invent impact to strengthen the wording.
Assessment Record
Record each answer with its evidence or policy source and uncertainty. Keep the following dimensions distinct:
| Dimension | Interpretation |
|---|---|
| Technical validity | Independent positive/control evidence supports, refutes, or leaves the claimed behavior unresolved. |
| Scope and eligibility | Approved asset/technique boundaries and supplied program criteria; missing policy remains uncertain. |
| Novelty | Exact overlap, supported variant, or unknown based on available prior reports. |
| Report quality | Reproduction prerequisites, evidence references, and accurate title/impact wording. |
A question-level record may use the existing gate_q1_scope, gate_q2_impact,
and gate_q3_poc_proves_impact properties, with corresponding entries for Q4–Q7.
Each entry needs a reason and source reference. Do not collapse the answers into
an all-or-nothing technical verdict or discard evidence for a business exclusion.
Common Assessment Errors
| Error | Consequence |
|---|---|
| Treat validator success as sufficient proof of impact | An oracle may establish only a narrower behavior. |
| Infer eligibility without the supplied policy | Technical validity becomes confused with program acceptance. |
| Treat unknown disclosure history as unique | Novelty is overstated without evidence. |
| Reject a technically proven issue for a duplicate or weak title | Valid evidence is lost for an unrelated disposition. |
| Apply evidence scrutiny only to High/Critical findings | Lower-severity claims can remain unsupported. |
Cross-references
- Verifier agent prompt:
decepticon/agents/prompts/verifier.md - Verifier router:
skills/verifier/SKILL.md - Bounty report skill:
skills/verifier/bounty-report/SKILL.md - Operator's external
triage-validationskill (Decepticon-external) for the broader gate methodology
