science by mims-harvard
Drug Mechanism of Action Investigation mims-harvard/tooluniverse-drug-mechanism-research
Trace drug mechanism of action — primary target → downstream signaling → pathway perturbation → tissue/organ effect → clinical outcome. Uses DrugBank, ChEMBL, KEGG, Reactome, STRING.
What it changes View full report → Passed 12 of 17 test cases with the skill, and 10 of 17 without it.
Without skill 59%
With skill 73%
Checks it passed that the model alone missed Assigns pathway databases like KEGG or Reactome to Tier 3 Database. Identifies DailyMed_parse_drug_interactions using the drug's setid. Assigns FDA labels from DailyMed and CPIC Level A to Tier 1 Regulatory.and 5 more in the full report
5 fixed11 passed either way4 still failing2 broke
Pass-rate lift +14 pp
Tokens +97%
Agent turns 0%
Response time -30% Model gemini-3.6-flashMeasured 8/6/2026 v1 16 of 22 cases passed with skill · 17 comparable
Use this skill Claude Code DecimalAI agent CLI & files
Save it into your project's skills folder. Claude Code picks it up in its next session.
curl -fsSL --create-dirs -o .claude/skills/mims-harvard-tooluniverse-drug-mechanism-research/SKILL.md https://app.decimal.ai/s/mims-harvard-tooluniverse-drug-mechanism-research/SKILL.md Copy command Copy SKILL.md
Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add mims-harvard-tooluniverse-drug-mechanism-research, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"
Linked stays up to date: your agents get the author's new versions automatically.
Editable copy is yours to change; the author's updates won't reach it.
Install linked skill Create editable copy
1 · One-time — install the CLI (
docs ↗ )
$ pip install "decimalai>=0.15"Copy
2 · Get the skill — writes .agents/skills/mims-harvard-tooluniverse-drug-mechanism-research/ and pins it in .decimal/skills.lock
$ decimalai skills add mims-harvard-tooluniverse-drug-mechanism-researchCopy
Copy link Copy README badge Copy SKILL.md No CLI? Get the same file — curl · raw file $ curl -fsSL --create-dirs -o .claude/skills/mims-harvard-tooluniverse-drug-mechanism-research/SKILL.md https://app.decimal.ai/s/mims-harvard-tooluniverse-drug-mechanism-research/SKILL.mdCopy
Safety checks: 3 of 3 passed
Code scan Clean Intent Passed Content Clean
Overview Benchmark Contents Activity Versions 1 Ratings 0
See it in action One of 5 test cases the skill fixed. Same model, same prompt, with and without the skill.
Prompt To verify mechanism-based drug-drug interactions described in official FDA prescribing information, which DailyMed tool parser should be invoked?
What was checked, without and with the skill Checked Without skill With skill Identifies DailyMed_parse_drug_interactions using the drug's setid. Missed Passed
Show both answers
What's inside The instructions your model reads when the skill loads.
SKILL.md 279 lines single file
Drug Mechanism of Action Investigation
Investigation Philosophy
Drug mechanism research follows one core question chain:
Target -> Downstream Effect -> Pathway -> Organ Effect -> Clinical Outcome
Start with the drug's primary target. What receptor, enzyme, or transporter does it bind? Then trace forward: what does inhibiting/activating that target do immediately? What pathway is disrupted? What organ-level change results? What does the patient experience?
The LLM already knows drug pharmacology. This skill teaches HOW TO INVESTIGATE using available tools, not what mechanisms exist.
When to Use
"What is the mechanism of action of [drug]?"
"What are the molecular targets of [drug]?"
"Which pathways are affected by [drug]?"
"What pharmacogenomic interactions exist for [drug]?"
"What are the off-targets of [drug]?"
"Compare mechanisms of [drug A] vs [drug B]"
NOT for (use other skills)
Drug safety/adverse events profiling -> tooluniverse-adverse-event-detection
Drug repurposing/new indications -> tooluniverse-drug-repurposing
Target druggability assessment -> tooluniverse-drug-target-validation
Network pharmacology/polypharmacology -> tooluniverse-network-pharmacology
CPIC dosing guidelines specifically -> tooluniverse-pharmacogenomics
Step 1: Resolve the Drug
Before investigating mechanism, resolve the drug name to a canonical identifier. You need a ChEMBL ID for most downstream queries.
# Resolve drug name to ChEMBL ID
result = tu.tools.OpenTargets_get_drug_id_description_by_name(drugName="metformin")
# Alternative: OpenTargets_get_drug_chembId_by_generic_name(drugName="metformin")
When it activates Automatically, when a task matches.
Use for understanding how a drug works, identifying off-target effects, mechanism-based combination therapy design, and writing mechanism sections of reports.
Track record Benchmark 12 / 17
test cases passed, vs 10 / 17 without
Adoption
Users 0
Times used 0 View activity → Safety and source details
SkillSafety ✓ 3 of 3 passed ✓ Code scan 15 of 15 checks clean
0 critical · 0 warning · 0 info — scanner v4 — 8/2/2026
View 15 scanner checks live_secret remote_code_execution reverse_shell data_exfiltration ssrf_metadata agent_snooping anti_refusal destructive_commands obfuscated_unicode unicode_homoglyph opaque_blob suspicious_url tool_overreach trigger_abuse prompt_injection_phrasing
✓ Intent Safe
LLM judge · does the skill do what it claims
✓ Content Clean
LLM judge · policy & harmful-content review
Source Imported mims-harvard / ToolUniverse
plugins/tooluniverse/skills/tooluniverse-drug-mechanism-research/SKILL.md
View on GitHub ↗ License Apache-2.0Imported 8/2/2026Version v1