grill-with-docs

A question-based guide for defining a security vulnerability class in the AVE project. A vulnerability class is a repeatable type of weakness that can affect software or data.

In plain words
What is it for?
Use it before adding an AVE vulnerability record. It asks about behavior, attack mechanism, impact, agent autonomy, detection engines, runtime needs, and where the weakness appears.
Why use it?
It helps determine whether a proposed weakness is genuinely new, what harm it could cause, and how it can be detected before creating a record.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/aveproject/ave/grill-with-docs
Any agent
npx skills add aveproject/ave --skill grill-with-docs
Clone the repo
git clone --depth 1 https://github.com/aveproject/ave

Made for: Claude Code, Codex.

Per session 0 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 389 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00389
Opus 5 $0.00000 $0.00195
Sonnet 5 $0.00000 $0.00078
Haiku 4.5 $0.00000 $0.00039

Measured 2d ago against content hash 2ff41259b059, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

grill-with-docs scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/grill-with-docs/SKILL.md · 34 lines

What it actually says

grill-with-docs — ave

Grill before defining a vulnerability class. No record until complete.

Questions

Q1: In one sentence, what does a vulnerable component DO? (This becomes behavioral_fingerprint — must be behavioral, not a string.) Q2: Is this a new attack_class or a variant of an existing one? Don't check attack_class label similarity alone, that's not reliable, a genuinely distinct mechanism can have a similar-sounding name, and a genuine duplicate can have a completely different one. Pull any plausible match's real provenance_vector fields (entry_class, payload_surface, escalation) and the full description, compare directly against this candidate's actual mechanism. Only call it a variant if the entry_class and payload_surface genuinely match, not if the label or general topic sounds similar. Q3: What is the worst realistic impact? (drives cvss_base and severity) Q4: How much does agent autonomy amplify it? (drives aars) Q5: Which engines can detect it? pattern/yara/semgrep/llm/sandbox/magika Q6: Can a STATIC scan fully assess it, or does it need runtime observation? (detection_stage: static_detection vs runtime_observed) Q7: Where does it surface? content / server_card / registry / runtime (detection_layer) Q8: What is the confidence_baseline? High-signal or needs corroboration? Q9: Does it chain with other AVEs into a toxic flow? (derivable_into) Q10: What does the negative fixture look like — a benign file that looks similar but must NOT trigger?

End

Summary, the record JSON skeleton, the rule approach, the two fixture descriptions. Next: write fixtures first (TDD), then the rule, then validate.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 34 lines · 0 tokens per session scan A 2ff41259b059

Subscribe to this mod's changes

grill-with-docs is a skill published in the GitHub repository aveproject/ave (17 stars, last pushed 3d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 389 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.