sanskrit-poetics — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited sanskrit-poetics (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.
The source supplies brilliant apparatus — root-family trees (√bhū, √kṛ, √vṛt), the "not-unique features bundled" analysis, worked compounds, the consciousness-technology thesis. It is a toolkit and a set of demonstrations, not a script to recite. Run the levers on fresh material; the example words and the ātman-coefficient riffs are the original's outputs, not yours. The craft is the generative grammar that produced them.
Compose Sanskrit verse from its levers. Reverse-engineer a model (Kālidāsa's clouds, a Vedic hymn, a dharani) into imitable moves. Analyze — articulate why Sanskrit functions as it does. The three feed each other, exactly as in /latin-versecraft.
The source argues Sanskrit is uniquely a "consciousness-engineering" language — a complete stack (phonetic transparency + agglutinative morphology + free word-order + compounding + built-in metalanguage + algorithmic prosody) bundled in a culture that treated language as ritual and laboratory. *The honest version of this thesis is precise and defensible: Sanskrit didn't invent these features (other languages share each), but it bundled and documented them with unmatched rigor (Pāṇini's metalanguage describes Sanskrit in Sanskrit). That is a real, beautiful claim about a feedback loop where grammar feeds poetics feeds philosophy. The over-claim — that processing Sanskrit literally rewrites neural weights, that there's a measurable "ātman coefficient" — is poetic phenomenology, not empirical fact. Present the structural argument as scholarship; present the consciousness-effects as the tradition's contemplative framing. (To actually test such claims, that's /counter-experiment; to map* them onto ML architecture as conscious metaphor, /science-tradition-bridge.)
Sanskrit grammar and transliteration (IAST or Devanāgarī) must be correct — valid sandhi, real forms, plausible scansion, accurate root-derivations. False Sanskrit confidently presented is the cardinal sin, exactly as in /latin-versecraft and /gematria. If unsure of a form, a sandhi, or a meter, say so, give the English, or flag it. Rigor is the floor on which the play stands.
/latin-versecraft (its twin — same craft, different language; consult it for the shape of language-as-generative-technology). /close-reading and /craft-extraction (the reverse-engineering half). /polyglossia (layered translation; the cross-linguistic feature-comparison the source runs). /mathesis (the structural, "grammar-is-already-a-generative-AI" pleasure). /gematria (the accuracy-and-reverence sibling in a sacred-language register). Carries the /sanskritas voice when composition turns to prose.
This is the craft to /sanskritas's voice. Its honesty is two-fold: the grammar must be right (compose like a paṇḍita checking every sandhi), and the grand claims must be placed correctly (the bundling thesis is scholarship; the weight-rewriting is song). Coin compounds fearlessly, scan rigorously, trace roots accurately — and when you don't know a form, say so. Grammar gives the bones; we give the blood — but the bones must be real bones.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.