Blockbuster Therapy Predictor
v0.1.0Predict which early-stage biotechnology platforms (PROTAC, mRNA, gene editing, etc.) have the highest potential to become blockbuster therapies. Analyzes cli...
⭐ 0· 102·0 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
medium confidencePurpose & Capability
The name/description promise multi-source aggregation (clinical trials, patents, funding). The included Python code implements scoring and contains built-in market estimates and scoring logic, but I see no declared connectors, required credentials, or network libraries in the visible code. This is a minor mismatch: the tool can perform local analyses but does not appear to automatically fetch external datasets as the documentation suggests.
Instruction Scope
SKILL.md instructs running scripts/main.py with options to output or save reports. It mentions optional access to clinical, patent, and funding databases but provides no concrete steps or env var names for connecting to those services. The runtime instructions do not direct reading unrelated system files or exfiltration; saving reports to disk is expected behavior.
Install Mechanism
No install spec is provided (instruction-only plus included script). requirements.txt only lists 'dataclasses' and 'enum', which are standard in Python 3.8+ (dataclasses is built-in in 3.8), so nothing suspicious is being fetched or installed by default.
Credentials
The skill declares no required environment variables or credentials, which is appropriate for a local scoring tool. However, SKILL.md's mention of optional access to external databases implies users might supply credentials; those are not enumerated here, so if you later configure connectors, expect to provide database/API credentials—verify how/where they are stored.
Persistence & Privilege
always is false and the skill does not request persistent agent privileges. The script only reads inputs and can write report files; there is no evidence it modifies other skills or system-wide agent settings.
Assessment
This package appears to be a local Python scoring tool (no credentials requested), but its documentation claims multi-source data aggregation that the included code does not visibly implement. Before installing or running with real data: 1) review the full scripts/main.py (the provided snippet was truncated) to confirm there are no network calls or hidden endpoints; 2) if you plan to connect it to clinical/patent/funding APIs, only supply credentials you control and prefer using a sandbox or limited-permission account; 3) note requirements.txt lists stdlib packages (dataclasses, enum) which is harmless but unnecessary for Python 3.8+; and 4) run the script in an isolated environment first and inspect any output files to ensure no sensitive data is being written or sent externally.Like a lobster shell, security has layers — review code before you run it.
latestvk976yxsh0rvxka3w73zw837nw58335wf
License
MIT-0
Free to use, modify, and redistribute. No attribution required.