Crispr Grna Designer
v0.1.0Design CRISPR gRNA sequences for specific gene exons with off-target prediction and efficiency scoring. Trigger when user needs gRNA design, CRISPR guide RNA...
⭐ 0· 120·0 current·0 all-time
byAIpoch@aipoch-ai
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (gRNA design, on-target/off-target scoring) match the provided files: SKILL.md describes Ensembl lookups, scoring algorithms, and off-target alignment; scripts/main.py implements sequence fetching (mocked or via requests), PAM finding, scoring, and simulated off-target checks. Declared dependencies (Biopython, pysam, numpy, pandas, requests) are typical for this domain.
Instruction Scope
Runtime instructions and SKILL.md direct the agent to fetch sequences from Ensembl (network), run local aligners (bowtie2/BWA) or pysam-based checks, and read/write results to the workspace. These actions are expected for gRNA design, but they grant filesystem and network access — so outputs and downloaded reference data should be handled in an isolated/sandboxed environment and validated before experimental use.
Install Mechanism
There is no automated install spec in the registry (instruction-only), lowering installation risk; however a requirements.txt is provided. Installing Python deps (pysam, biopython) and optional native tools (bowtie2, ViennaRNA) may require compiling native code or installing system packages. No remote arbitrary archive downloads are present in the metadata.
Credentials
The skill does not request environment variables, credentials, or config paths. All network access is to public reference services (Ensembl is referenced) and optional local tools. No apparent need for unrelated secrets or cloud credentials.
Persistence & Privilege
Registry flags are default: not always-included, user-invocable, and allows model invocation (normal). The skill does not request to modify other skills or system-wide settings. It writes to workspace per SKILL.md, which is expected for a tool that outputs results.
Assessment
This skill appears coherent for designing CRISPR gRNAs, but it operates in a high-risk technical domain. Before installing or running: 1) Review the scripts/main.py source yourself (or have a trusted reviewer) — the package comes from an unknown source. 2) Install and run in an isolated sandbox or VM (not on production systems) because it downloads/reference-checks sequences and may invoke native aligners. 3) Audit and pin Python dependencies (note the ambiguous 'bio' entry) and be cautious installing pysam or other native-built packages. 4) Expect network calls to Ensembl (or mocked data if requests is absent); confirm no unexpected external endpoints are used. 5) Do not use outputs directly for lab experiments without independent validation — SKILL.md explicitly warns predictions need experimental confirmation. 6) If you are concerned about autonomous agent actions, restrict usage to user-invoked only or disable autonomous execution. 7) Ensure compliance with your institution's biosecurity and ethics policies before designing or using guides generated by this tool.Like a lobster shell, security has layers — review code before you run it.
latestvk979mdtnp4x1zf618j97wpj5as834r38
License
MIT-0
Free to use, modify, and redistribute. No attribution required.