{"id":"cmuhnnezlm7fqad075ap9vl29","slug":"axolotl","publicId":"08171b1f-7d1c-44f1-b466-ddc437a9eea6","name":"axolotl","description":"Axolotl: YAML LLM fine-tuning (LoRA, DPO, GRPO).","version":"1.0.0","author":"Orchestra Research","license":"MIT","compatibility":"","status":"completed","overallGrade":"B","scannerVersion":9,"fileCount":5,"hasSkillMd":true,"hasScripts":false,"hasReferences":true,"hasEvals":false,"hasAssets":false,"sourceType":"github","sourceUrl":"https://github.com/NousResearch/hermes-agent/tree/main/optional-skills/mlops/training/axolotl","createdAt":"2026-09-26T00:32:08.865Z","completedAt":"2026-09-26T00:32:08.862Z","visibility":"public","publishedAt":"2026-06-08T14:31:26.932Z","category":"","frameworks":[],"badgeStatus":"conditional","downloadCount":0,"verdictRationale":{"summary":"Skill has warnings: 2 warnings in security or structure checks.","verdict":"warn","scanners":[{"source":"vettd","weight":1,"verdict":"warn"},{"source":"cisco","weight":0.8,"verdict":"pass"}],"consensus":"warn","warnTriggers":["Embedded chat-template special token","References external URL — review for indirect prompt injection risk"],"policyVersion":"consensus-v1","hardFailTriggers":[]},"verdictPolicyVersion":"consensus-v1","skillId":"cmq5b726h00buad062dir7ha2","findings":[{"id":"cmuhnnezmm7g5ad07hpms1l9h","source":"vettd","category":"security","label":"Embedded chat-template special token","detail":"Detected in references/dataset-formats.md:142 (9 matches) — `by using delimiters (<|im_start|> and <|im_end|>), a prompt separates different speakers which helps the model identify `","severity":"medium","ruleId":"VTD-0076","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"3f00fcfdd53f3052c53d58d8abf9535b","filepath":"references/dataset-formats.md"},{"id":"cmuhnnezmm7g7ad074g4vtro0","source":"vettd","category":"security","label":"References external URL — review for indirect prompt injection risk","detail":"External URL(s) detected in SKILL.md — referenced content can change after audit","severity":"medium","ruleId":"VTD-0088","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"42638a8b7e9fc5df27449992eaebc724","filepath":"SKILL.md"},{"id":"cmuhnnezmm7g8ad0786b5a075","source":"cisco","category":"security","label":"Dangerous data flow in command pipeline","detail":"references/dataset-formats.md:1 — Pipeline reads data and passes it to code execution: `head -n1 output.jsonl | python -m json.tool`. Review for potential command injection. (Note: found in documentation file - may be instructional rather than executable.)","severity":"low","ruleId":"PIPELINE_TAINT_FLOW","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"da776c621c8cad4ace79be0778aeeb34","filepath":"references/dataset-formats.md"},{"id":"cmuhnnezmm7g9ad07026v9fwa","source":"cisco","category":"security","label":"Dangerous data flow in command pipeline","detail":"references/other.md:2 — Pipeline downloads data from the network and executes it: `curl -LsSf https://astral.sh/uv/install.sh | sh`. This is a remote code execution pattern. (Note: uses a well-known installer URL - likely a standard installation command.)","severity":"low","ruleId":"PIPELINE_TAINT_FLOW","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"c35492c9067a6c175d8545be08a8d011","filepath":"references/other.md"},{"id":"cmuhnnezmm7g6ad07xiozifjm","source":"vettd","category":"security","label":"No prompt injection or jailbreak signals detected","detail":"Scanned text content for instruction override, jailbreak framing, credential solicitation, and embedded injection markers","severity":"info","ruleId":"VTD-0092","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"ee5675c6452295fc92108d4c30cfebf6","filepath":""},{"id":"cmuhnnezlm7fwad07lf6gewp3","source":"vettd","category":"security","label":"No repository link","detail":"No repository field found in SKILL.md frontmatter. Skills without a verifiable source repository cannot be externally audited.","severity":"info","ruleId":"VTD-0083","owaspLlmCategory":null,"intent":"negligent","chainId":null,"fingerprint":"e09767e7943872a2754f559ba01ef7db","filepath":""},{"id":"cmuhnnezmm7g4ad07cax16rtc","source":"vettd","category":"security","label":"No secrets or unsafe code patterns detected","detail":"Scanned all files for credentials, private keys, and code-level risks (eval, shell exec, destructive ops)","severity":"info","ruleId":"VTD-0091","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"2e3428ecd3c04cac33cabd959f51437a","filepath":""},{"id":"cmuhnnezlm7fsad07uubuv2bl","source":"vettd","category":"structure","label":"No scripts/ directory","detail":"Consider bundling reusable scripts for validation and automation","severity":"info","ruleId":"VTD-0096","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"21a167b473d29b0a653734953d43b2fc","filepath":""},{"id":"cmuhnnezlm7ftad07hgwioqp6","source":"vettd","category":"structure","label":"references/ directory present","detail":"Additional documentation files available for progressive disclosure","severity":"info","ruleId":"VTD-0097","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"a5f22b7185c58877c3d5a94338930d62","filepath":""},{"id":"cmuhnnezlm7frad071qmthjjq","source":"vettd","category":"structure","label":"SKILL.md present","detail":"Required skill definition file found","severity":"info","ruleId":"VTD-0095","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"c4ab14319b89c3d0802fadf0dd105bac","filepath":""},{"id":"cmuhnnezlm7fvad07iasyofju","source":"vettd","category":"structure","label":"Valid name field","detail":"Name \"axolotl\" follows spec (lowercase, hyphens, ≤64 chars)","severity":"info","ruleId":"VTD-0099","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"5ba54eca0b3e299a99112de98cbcd97f","filepath":""},{"id":"cmuhnnezmm7fyad07fjdlfb6u","source":"vettd","category":"description","label":"Description lacks usage context","detail":"Add context like \"Use this skill when...\" to help agents know when to activate it","severity":"info","ruleId":"VTD-0111","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"e325ab773f2c18f6f9d864add6464110","filepath":""},{"id":"cmuhnnezlm7fxad0730rze50m","source":"vettd","category":"description","label":"Description within character limit","detail":"48/1024 characters used","severity":"info","ruleId":"VTD-0110","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"66553d595a652c45a6f0d4353184b37e","filepath":""},{"id":"cmuhnnezmm7g0ad07ocn6qmeq","source":"vettd","category":"best-practices","label":"Examples included","detail":"Found code blocks, input/output samples, or an examples section — concrete samples help agents pattern-match effectively","severity":"info","ruleId":"VTD-0103","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"c2ceb48c4af3c87dafa32ec2596251b8","filepath":""},{"id":"cmuhnnezmm7g3ad079dd139sc","source":"vettd","category":"best-practices","label":"Progressive disclosure used","detail":"SKILL.md body references files in references/, scripts/, or assets/ — agents can load additional context on demand instead of consuming everything upfront","severity":"info","ruleId":"VTD-0107","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"6fd5870d64ef20cb2bbaf421f8655a32","filepath":""},{"id":"cmuhnnezmm7fzad071ceinq20","source":"vettd","category":"best-practices","label":"SKILL.md body length is reasonable","detail":"153 lines (recommended: under 500)","severity":"info","ruleId":"VTD-0101","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"42debaa072535bf5650f90d3168f0b93","filepath":""},{"id":"cmuhnnezmm7g1ad07pj4nisfo","source":"vettd","category":"best-practices","label":"Step-by-step workflow found","detail":"Structured procedures improve reliability for complex tasks","severity":"info","ruleId":"VTD-0106","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"2368c9a31ac52269c854f105795cf28c","filepath":""},{"id":"cmuhnnezmm7g2ad0709hdbgx8","source":"vettd","category":"best-practices","label":"Validation loop referenced","detail":"Instructions for the agent to validate its own work before proceeding","severity":"info","ruleId":"VTD-0105","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"e5a31b775d1df438cbeb1e3fcb2ff503","filepath":""},{"id":"cmuhnnezlm7fuad07f1juykgj","source":"vettd","category":"evals","label":"No evaluation test cases","detail":"Add an evals/ directory with test prompts and expected outputs to measure skill quality","severity":"info","ruleId":"VTD-0118","owaspLlmCategory":null,"intent":null,"chainId":null,"fingerprint":"41d019e0494a4a9763fc23c948d14e09","filepath":""}],"scannerRuns":[{"source":"cisco","status":"success","verdict":null,"grade":"A","findingCount":2,"criticalCount":0,"highCount":0},{"source":"vettd","status":"success","verdict":null,"grade":"B","findingCount":17,"criticalCount":0,"highCount":0}],"user":{"name":"Justin Pacella","image":"https://avatars.githubusercontent.com/u/97124458?v=4"},"repository":{"owner":"NousResearch","name":"hermes-agent"},"freshness":null,"signalCategories":[{"category":"safety","label":"Safety","form":"graded","verdict":null,"rows":[]},{"category":"reliability","label":"Reliability & Quality","form":"graded","verdict":null,"rows":[]},{"category":"performance","label":"Performance","form":"measured","verdict":null,"rows":[]},{"category":"cost","label":"Cost","form":"measured","verdict":null,"rows":[]},{"category":"compatibility","label":"Compatibility","form":"unjudged","verdict":null,"rows":[]},{"category":"sentiment","label":"Popularity","form":"unjudged","verdict":null,"rows":[]},{"category":"characteristics","label":"Characteristics","form":"unjudged","verdict":null,"rows":[]}],"signalCount":25,"publisherSupplied":["name","description","author","version","license"]}