doc-summariser — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited doc-summariser (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
An agent skill that demonstrates the SKILL.md + helper-code pattern with a real, runnable capability.
Given a block of text, produce a faithful summary. Two modes are available:
summarise_document): selects the most salient wholesentences from the text using a frequency-based scoring scheme (a variant of the classic Luhn method) and returns them in their original order, so the summary reads naturally.
gist): a word-bounded single line for when the callerwants the shortest possible version.
Both modes are deterministic and run fully offline, which makes the skill cheap, reproducible, and easy to test. They share their implementation with the MCP summarise tool, so there is one source of truth to upgrade later (for example, to a Claude-generated abstractive summary).
version".
from agent_skills.doc_summariser.helper import summarise_document, gist
# Most salient sentences, original order:
summarise_document(long_text, max_sentences=3)
# A single short line:
gist(long_text, max_words=20)Or run the skill directly to see both modes on a sample:
python agent_skills/doc_summariser/helper.pytext (str): the document to summarise.max_sentences (int, default 3): upper bound on extractive summary length.max_words (int, default 20): upper bound on the one-line gist.summarise_document: a string containing the top sentences, joined and inoriginal document order. If the text is already within the sentence bound it is returned (whitespace-normalised) unchanged.
gist: a single short line; the original text if already short enough,otherwise the first max_words words followed by an ellipsis.
ValueError.~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.