Methodology inspired by publicly taught pipeline-valuation frameworks; all text is an original paraphrase.
Defaults
| Parameter |
Default Value |
Rationale |
| asset_scope |
all disclosed assets |
Full-pipeline enumeration first |
| pos_source |
published phase transition ranges |
Starting point, then adjusted with logged reasons |
| status_diff |
true |
Previous vs current trial status shown per asset |
| competitor_map |
true |
Same-target and same-indication crowding surfaced |
| value_frame |
risk-adjusted |
Modeled value = unadjusted peak x POS, both shown |
Preflight
Run canonical pre-flight per contracts/preflight.md. Include the X-Agentii-Trace header on every tool call per contracts/x-agentii-trace-header.md — carry the _run_id from your first tool result and name yourself (and your parent, if you were spawned). Confirm ticker resolution via search_companies before asset queries.
Triggers
- "Analyze [ticker]'s drug pipeline by phase."
- "What are the Phase II and III assets for [ticker]?"
- "Risk-adjust [ticker]'s pipeline with explicit POS assumptions."
- "Which of [ticker]'s programs is most valuable?"
- "What trials recently changed status for [ticker]?"
- "Who else is targeting [mechanism/target]?"
- "How crowded is [indication], and what does that mean for [ticker]?"
- "Log every POS change with reasons for [ticker]'s pipeline."
- "What is the next catalyst for each of [ticker]'s assets?"
- "Compare [ticker]'s pipeline against same-indication competitors."
- "Map [ticker]'s assets by mechanism of action."
- "What does [ticker]'s pipeline say about its runway?"
Production Grounding
- POS discipline: every asset carries an unadjusted peak sales estimate and a phase-appropriate POS; the modeled number is peak x POS, never a bare risk-adjusted figure. Every POS change is logged with its reason (trial data, competitor data, discontinuation) — silent POS drift is a defect.
- Phase ladder: Phase I safety/dosing, Phase II efficacy signal, Phase III registrational; option value rises with phase. Apply scrutiny axes when sizing pivotal assets.
- Status-diff discipline: show previous vs current status per trial (New/Suspended/Terminated/Ahead/Delayed/...); Terminated plus enrollment collapse is a red flag; recruitment-complete starts the readout clock.
- Competitor mapping: same-target and same-indication crowding discounts peak sales, not just POS; mechanism overlap explains class risk.
- Grounding detail lives in
references/knowledge-frameworks.md.
Data Source Priority
get_company_drugs / search_universe_drugs — asset inventory from the med universe.
search_drug_knowledge / get_drug_knowledge — mechanism, targets, indications, phase per drug.
search_drugs_by_target / search_drugs_by_indication — reverse lookups for the competitor map.
search_clinical_trials / get_clinical_trial — status, enrollment, primary-completion dates.
search_fda_approvals — registration history; search_earnings_calendar for timing; knowledge layer via search_investment_cases / search_knowledge_entries.
Methodology
Retrieval Scope
structured_only
Retrieval Strategy
- Pull asset inventory (
get_company_drugs), then universe cross-checks (search_universe_drugs).
- Enrich per asset from
search_drug_knowledge / get_drug_knowledge (mechanism, targets, phase).
- Pull trial records (
search_clinical_trials) for status-diff and next-catalyst dates.
- Build the competitor map via
search_drugs_by_target / search_drugs_by_indication.
- Size each asset (peak x POS with change log); ground in cases and knowledge entries.
Temporal Scope
See frontmatter temporal_scope block. Development programs span years; status history may reach back 8-12 quarters.
Tool Allowlist
See frontmatter allowed_tools.
Protocol
- Asset enumeration
- Phase and mechanism enrichment
- Status-diff and catalyst dating
- Competitor map assembly
- POS model with change log
Modes
- Phase scan (default): all assets by phase with status-diff and next catalyst.
- POS model: peak x POS modeling with a full change log and scenario brackets.
- Competitor map: target/indication crowding with class-risk read.
Tool Fallbacks
| Failure |
Fallback |
| get_company_drugs empty |
search_universe_drugs by company or indication; annotate coverage_gap |
| Drug-knowledge record missing |
Flag mechanism/competitor data unavailable; never infer targets |
| No trial rows |
Mark status undisclosed; note the readout clock is unverified |
| Knowledge tools empty |
Proceed with structured data only |
Output File
{ticker}/{YYYY-MM-DD_HHMM}_pipeline-medicines_{affix}.md
Output Structure
- Executive Summary — pipeline stance in 2-3 sentences
- Asset Table — phase, indication, mechanism, status-diff, next catalyst
- POS Model — unadjusted peak, POS, modeled value per asset, change log with reasons
- Competitor Map — same-target and same-indication crowding, class risk
- Historical Context — cases and knowledge entries with /v/ citations
- Coverage Gaps — missing records, undisclosed status, degraded modes
Error Handling
| Error |
Fallback |
| Phase unknown |
Mark "undisclosed"; do not guess |
| POS source missing |
State the assumption range and label it a judgment |
| No catalysts found |
Say so explicitly; note the pipeline may be early-stage |
Memory Load
See contracts/memory-load.md.
Snapshot
See contracts/snapshot-synthesis.md.
Final Summary (TUI)
Include ### Key Citations block with 0-10 clickable /v/ URLs.
References
contracts/citation-and-memory.md
contracts/output-frontmatter-schema.md
contracts/memory-load.md
contracts/snapshot-synthesis.md
contracts/preflight.md
references/knowledge-frameworks.md
1---2name: pipeline-medicines3description: "Medicines pipeline analysis over structured universes: Phase I-III assets with probability-of-success discipline (unadjusted peak x POS = modeled value; POS changes logged with reasons), trial status-diff tracking, and mechanism/competitor mapping via drug-knowledge lookups."4sectors: [med.medicines_biotech, med.medical_devices]5multi_ticker_semantics: single_target6temporal_scope:7 default_quarters: 88 max_quarters: 209 description: "Long-horizon pipeline window default 8 quarters; up to 20 for multi-year development programs."10allowed_tools:11 - get_company_drugs12 - search_universe_drugs13 - search_drug_knowledge14 - get_drug_knowledge15 - search_drugs_by_target16 - search_drugs_by_indication17 - search_clinical_trials18 - get_clinical_trial19 - search_fda_approvals20 - search_companies21 - get_company_profile22 - search_earnings_calendar23 - search_investment_cases24 - search_knowledge_entries25retrieval_scope: structured_only26min_tool_diversity: 327parameter_free: false28metadata:29 source: https://skillmd.com/skills/agentii-ai/pipeline-medicines30---3132> Methodology inspired by publicly taught pipeline-valuation frameworks; all text is an original paraphrase.3334## Defaults3536| Parameter | Default Value | Rationale |37|-----------|---------------|-----------|38| asset_scope | all disclosed assets | Full-pipeline enumeration first |39| pos_source | published phase transition ranges | Starting point, then adjusted with logged reasons |40| status_diff | true | Previous vs current trial status shown per asset |41| competitor_map | true | Same-target and same-indication crowding surfaced |42| value_frame | risk-adjusted | Modeled value = unadjusted peak x POS, both shown |4344## Preflight4546Run canonical pre-flight per `contracts/preflight.md`. Include the `X-Agentii-Trace` header on every tool call per `contracts/x-agentii-trace-header.md` — carry the `_run_id` from your first tool result and name yourself (and your parent, if you were spawned). Confirm ticker resolution via `search_companies` before asset queries.4748## Triggers4950- "Analyze [ticker]'s drug pipeline by phase."51- "What are the Phase II and III assets for [ticker]?"52- "Risk-adjust [ticker]'s pipeline with explicit POS assumptions."53- "Which of [ticker]'s programs is most valuable?"54- "What trials recently changed status for [ticker]?"55- "Who else is targeting [mechanism/target]?"56- "How crowded is [indication], and what does that mean for [ticker]?"57- "Log every POS change with reasons for [ticker]'s pipeline."58- "What is the next catalyst for each of [ticker]'s assets?"59- "Compare [ticker]'s pipeline against same-indication competitors."60- "Map [ticker]'s assets by mechanism of action."61- "What does [ticker]'s pipeline say about its runway?"6263## Production Grounding6465- POS discipline: every asset carries an unadjusted peak sales estimate and a phase-appropriate POS; the modeled number is peak x POS, never a bare risk-adjusted figure. Every POS change is logged with its reason (trial data, competitor data, discontinuation) — silent POS drift is a defect.66- Phase ladder: Phase I safety/dosing, Phase II efficacy signal, Phase III registrational; option value rises with phase. Apply scrutiny axes when sizing pivotal assets.67- Status-diff discipline: show previous vs current status per trial (New/Suspended/Terminated/Ahead/Delayed/...); Terminated plus enrollment collapse is a red flag; recruitment-complete starts the readout clock.68- Competitor mapping: same-target and same-indication crowding discounts peak sales, not just POS; mechanism overlap explains class risk.69- Grounding detail lives in `references/knowledge-frameworks.md`.7071## Data Source Priority72731. `get_company_drugs` / `search_universe_drugs` — asset inventory from the med universe.742. `search_drug_knowledge` / `get_drug_knowledge` — mechanism, targets, indications, phase per drug.753. `search_drugs_by_target` / `search_drugs_by_indication` — reverse lookups for the competitor map.764. `search_clinical_trials` / `get_clinical_trial` — status, enrollment, primary-completion dates.775. `search_fda_approvals` — registration history; `search_earnings_calendar` for timing; knowledge layer via `search_investment_cases` / `search_knowledge_entries`.7879## Methodology8081### Retrieval Scope82structured_only8384### Retrieval Strategy851. Pull asset inventory (`get_company_drugs`), then universe cross-checks (`search_universe_drugs`).862. Enrich per asset from `search_drug_knowledge` / `get_drug_knowledge` (mechanism, targets, phase).873. Pull trial records (`search_clinical_trials`) for status-diff and next-catalyst dates.884. Build the competitor map via `search_drugs_by_target` / `search_drugs_by_indication`.895. Size each asset (peak x POS with change log); ground in cases and knowledge entries.9091### Temporal Scope92See frontmatter temporal_scope block. Development programs span years; status history may reach back 8-12 quarters.9394### Tool Allowlist95See frontmatter allowed_tools.9697### Protocol981. Asset enumeration992. Phase and mechanism enrichment1003. Status-diff and catalyst dating1014. Competitor map assembly1025. POS model with change log103104## Modes105106- **Phase scan** (default): all assets by phase with status-diff and next catalyst.107- **POS model**: peak x POS modeling with a full change log and scenario brackets.108- **Competitor map**: target/indication crowding with class-risk read.109110## Tool Fallbacks111112| Failure | Fallback |113|---------|----------|114| get_company_drugs empty | `search_universe_drugs` by company or indication; annotate coverage_gap |115| Drug-knowledge record missing | Flag mechanism/competitor data unavailable; never infer targets |116| No trial rows | Mark status undisclosed; note the readout clock is unverified |117| Knowledge tools empty | Proceed with structured data only |118119## Output File120121`{ticker}/{YYYY-MM-DD_HHMM}_pipeline-medicines_{affix}.md`122123## Output Structure1241251. **Executive Summary** — pipeline stance in 2-3 sentences1262. **Asset Table** — phase, indication, mechanism, status-diff, next catalyst1273. **POS Model** — unadjusted peak, POS, modeled value per asset, change log with reasons1284. **Competitor Map** — same-target and same-indication crowding, class risk1295. **Historical Context** — cases and knowledge entries with /v/ citations1306. **Coverage Gaps** — missing records, undisclosed status, degraded modes131132## Error Handling133134| Error | Fallback |135|-------|----------|136| Phase unknown | Mark "undisclosed"; do not guess |137| POS source missing | State the assumption range and label it a judgment |138| No catalysts found | Say so explicitly; note the pipeline may be early-stage |139140## Memory Load141142See `contracts/memory-load.md`.143144## Snapshot145146See `contracts/snapshot-synthesis.md`.147148## Final Summary (TUI)149150Include ### Key Citations block with 0-10 clickable /v/ URLs.151152## References153154- `contracts/citation-and-memory.md`155- `contracts/output-frontmatter-schema.md`156- `contracts/memory-load.md`157- `contracts/snapshot-synthesis.md`158- `contracts/preflight.md`159- `references/knowledge-frameworks.md`