BC operator-approval-loop
Operator approval contract with internal filing notices for agent-drafted outbound messages, hashed drafts, epoch-keyed decisions, durable delivery claims and receipts, and a pre-draft baseline gate. Use when an agent drafts messages to external counterparties and a human operator must approve, reject, or steer each send before it leaves.
Operator approval contract with internal filing notices for agent-drafted outbound messages, hashed drafts, epoch-keyed decisions, durable delivery claims and…
As a process C 51/100 · Has gaps — weak spots: result and completion, when it triggers, inputs and preconditions
How to improve
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 3. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- note
edit-residuethe text marks something as outdated (lines 93, 146): check that old rules are not kept next to new ones — the full check reads the text for contradictions
Process rating: all ten parameters 51/100
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 50Failures and branches. 0 branches, has a failure section
- 60Tools and files. Uses tools (bash, python) that frontmatter does not declare
- 100Steps. 28 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 3183 tokens
- 100Running it twice. Mutating operations check current state
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +3Description length 340: enough signal without eating the budget
- +4Structure: 16 headings
- +3Step-by-step instructions: 28 items
- +4Has examples (3 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 84.
In the sandbox The scripts kept to themselves
The skill's scripts were run in a throwaway machine: no network, fake keys in the home directory, a tracer watching. We wrote down what they did. Reaching for the network or for secrets caps the technical grade at C; a quiet run adds no points.
Запущено 1 скрипт; каждому дали двадцать секунд, поддельный домашний каталог с ключами и сеть, в которой ничего нет.
references/approval_claims.py | ничего за пределами своей папки |
6 Oct 2026