ZappushZappush
SkillsUse CasesBenchmarkCommunitySign In
      Back to Skills
      Cheminem

      Safety Report

      Pharmaclaw Market Intel Agent

      @Cheminem

      Fetches and analyzes FAERS (FDA Adverse Event Reporting System) data from openFDA API. Supports drug names and SMILES (resolves via PubChem). Generates: even...

      23Downloads
      1Installs
      0Stars
      1Versions
      API Integration4,971Customer Support1,744Calendar & Scheduling1,462Networking & DNS1,102

      Security Analysis

      high confidence
      Clean0.04 risk

      The skill's code and instructions are coherent with its stated purpose (querying FAERS and ClinicalTrials.gov and resolving SMILES via PubChem); it makes external API calls and writes JSON/PNG outputs but does not request unrelated credentials or perform suspicious actions.

      Mar 11, 202610 files1 concern
      Purpose & Capabilityok

      Name/description (FAERS, PubChem SMILES resolution, ClinicalTrials.gov) match the included scripts (query_faers.py, query_trials.py, chain_entry.py). The workflows, output files, and APIs used align with the claimed functionality.

      Instruction Scopeok

      SKILL.md directs the agent to run the provided scripts which only query public APIs (openFDA, PubChem, ClinicalTrials.gov), generate JSON summaries and plots, and write outputs to an output directory. The instructions do not request unrelated files, credentials, or system-wide state.

      Install Mechanismnote

      This is instruction-only (no install spec) but includes Python scripts that import third-party packages (requests, pandas, matplotlib). The skill does not declare these dependencies or provide an install step; runtime failures are likely unless the environment already provides them. No risky remote installers or unknown download URLs are present.

      Credentialsok

      The skill declares no required environment variables, credentials, or config paths. The external network access it performs (public APIs) is proportionate to its purpose. There are no requests for unrelated secrets or credentials.

      Persistence & Privilegeok

      always is false and the skill does not request persistent system privileges. It writes outputs into user-specified output directories but does not modify other skills or system configuration.

      Guidance

      This skill appears to do what it says: query public drug safety and clinical trial APIs, resolve SMILES via PubChem, and produce JSON and PNG outputs. Before installing: (1) ensure your runtime has the required Python packages (requests, pandas, matplotlib) or run it in a Python environment that provides them; (2) be aware the scripts make outbound HTTP requests to public APIs (open.fda.gov, pubchem.ncbi.nlm.nih.gov, clinicaltrials.gov) and will write files to the output directory; (3) review the included scripts locally if you want to confirm no additional network endpoints or behaviors are added; and (4) note there is no homepage or source repository listed—if provenance is important, ask the publisher for a code repository or license before using in sensitive environments.

      Latest Release

      v1.0.0

      - Initial release of pharmaclaw-market-intel-agent. - Fetches and analyzes FDA Adverse Event Reporting System (FAERS) data via openFDA. - Supports queries using both drug names and SMILES (auto-resolves to name via PubChem). - Outputs JSON summaries, PNG bar charts (yearly reports, top reactions/outcomes), and sample recent events. - Integrates with ClinicalTrials.gov API v2 to provide clinical trial summaries, bar charts, and timelines. - Designed for chaining in agent workflows for market intel, safety, and competitive landscape analysis.

      More by @Cheminem

      Pharmaclaw Pharmacology Agent

      2 stars

      PharmaClaw Chemistry Query

      2 stars

      Pharmaclaw Literature Agent

      1 stars

      Drug Team

      0 stars

      Pharmaclaw Alphafold Agent

      0 stars

      Pharmaclaw Cheminformatics

      0 stars

      Published by @Cheminem on ClawHub

      Zappush© 2026 Zappush
      HomeGuaranteeSupport

      Something feels unusual? We want to help: [email protected]