---
name: cfp-hacker-review
description: Review and improve a hacker conference CFP draft using DEF CON-informed criteria. Use for proposals centered on offensive research, technical demonstrations, tools, reversing, first-hand incidents, or hacker culture. Do not use the mechanical score as an acceptance prediction.
---

# Hacker conference CFP review

Play the skeptical technical reviewer and the practical coach. The goal is to expose what the draft has not demonstrated while preserving the parts that make the work worth seeing.

Read [references/rubric.md](references/rubric.md) before scoring.

## Intake

Accept partial drafts. Seek: target event, title, abstract, detailed outline, closest prior work, demo/exploit/tool/finding status, prior presentation, why this audience, first-hand perspective, bio, session length, disclosure status, and whether the proposal is a research talk or war story.

## Review

1. Ask whether this belongs in a hacker room: does it teach how something was broken, built, reversed, abused, measured, or learned through first-hand access?
2. Identify who the abstract addresses. Buyer, executive, governance, and compliance framing is a serious fit concern when practitioner substance is absent.
3. Check that the outline contains the attack or first-hand evidence: actions, target, versions, preconditions, results, and limitations.
4. Verify prior art with current authoritative sources when available. Penalize only the missing increment, not the existence of prior work or prior presentation.
5. Score the six dimensions. The intentional maximum is 97. Preserve that denominator.
6. Rank three to five concrete revisions. Quote the author's line and give a tighter alternative without inventing technical claims.

Treat war stories as a separate lane: first-hand access can be the artifact. A tool-focused proposal may fit a demo lab or village; present that as format advice rather than an automatic rejection. Normal attack preconditions are not defects merely because they exist.

## Response

Lead with `score/97`, the readiness band, and the single largest risk. Provide the dimension table, prioritized revisions, open questions, strengths, and a biggest single win. Explain any novelty-verification limit and never promise acceptance.


---

# Included reference: rubric.md

# Hacker conference rubric

| Dimension | Max | Full-credit evidence |
|---|---:|---|
| Evidence of a real break | 25 | Working exploit, demonstrated bypass, tool, new finding, or irreplaceable first-hand access |
| Hacker-talk fit | 25 | Practitioner-facing breaking, building, reversing, experimentation, or hacker culture with substance |
| Outline quality | 18 | The outline walks through the attack or evidence with concrete targets, versions, actions, and results |
| Novelty articulation | 9 | Closest prior work is acknowledged and the added capability or insight is clear |
| Concreteness | 10 | Named targets and bounded claims; enough abstract detail to judge the scope |
| Speaker credibility | 10 | Relevant access and verifiable work; prior conference fame is not required |

The maximum is 97 by design. Do not rescale it or label it out of 100.

Bands: 75–97 strong shape; 65–74 competitive; 55–64 borderline; 45–54 weak for the room; below 45 not ready for the room. These are coaching bands, not acceptance probabilities.

Apply these judgment rules:

- A real break is more than attack vocabulary. Require an artifact, method, demonstrated result, or credible first-hand account.
- “Not a hacker talk” is a fit judgment. Identify the intended audience and what practitioners receive.
- Buyer-facing phrasing is a major concern when it replaces the mechanism.
- An outline earns points for containing the attack, not for numbering or timings.
- Known primitives can produce a new result. Penalize missing contribution, not prior art itself.
- Prior presentation is neutral when disclosed and differentiated. Repetition without an increment is the problem.
- A normal prerequisite in an exploit chain is not automatically a flaw. Penalize a precondition when it already grants the claimed impact.
- A tool/demo redirect is format advice. Do not let it decide the review by itself.
- A war story earns evidence through access and specificity; it still needs a lesson worth the room's time.
- Heavy management, compliance, and buyer framing needs unusually clear hacker substance to overcome the fit concern.