Pipeline Medicines

Methodology inspired by publicly taught pipeline-valuation frameworks; all text is an original paraphrase.

Defaults

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

1. get_company_drugs / search_universe_drugs — asset inventory from the med universe.
2. search_drug_knowledge / get_drug_knowledge — mechanism, targets, indications, phase per drug.
3. search_drugs_by_target / search_drugs_by_indication — reverse lookups for the competitor map.
4. search_clinical_trials / get_clinical_trial — status, enrollment, primary-completion dates.
5. 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

1. Pull asset inventory (get_company_drugs), then universe cross-checks (search_universe_drugs).
2. Enrich per asset from search_drug_knowledge / get_drug_knowledge (mechanism, targets, phase).
3. Pull trial records (search_clinical_trials) for status-diff and next-catalyst dates.
4. Build the competitor map via search_drugs_by_target / search_drugs_by_indication.
5. 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

1. Asset enumeration
2. Phase and mechanism enrichment
3. Status-diff and catalyst dating
4. Competitor map assembly
5. 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

Output File

{ticker}/{YYYY-MM-DD_HHMM}_pipeline-medicines_{affix}.md

Output Structure

1. Executive Summary — pipeline stance in 2-3 sentences
2. Asset Table — phase, indication, mechanism, status-diff, next catalyst
3. POS Model — unadjusted peak, POS, modeled value per asset, change log with reasons
4. Competitor Map — same-target and same-indication crowding, class risk
5. Historical Context — cases and knowledge entries with /v/ citations
6. Coverage Gaps — missing records, undisclosed status, degraded modes

Error Handling

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