dcf-model-builder
Use when building public-equity DCF valuation workbooks. Default to the banker formula workbook path for new model builds; use deterministic exports only for controlled support calculations or explicit lightweight runs. Do not use for standalone workbook audits; use model-audit-tieout.
- License 许可
- MIT
- 最近更新
- 2026-09-22
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- dcf-model-builder
- description
- Use when building public-equity DCF valuation workbooks. Default to the banker formula workbook path for new model builds; use deterministic exports only for controlled support calculations or explicit lightweight runs. Do not use for standalone workbook audits; use model-audit-tieout.
02 资源文件 27
- 参考文档agents/openai.yaml285 B
- 资产文件assets/plan_template.json6.7 KB
- 资产文件assets/templates/banker_formula_workbook_template.xlsx36.1 KB
- 参考文档references/banker-formula-workbook-contract.md4.0 KB
- 参考文档references/cash-flow-methods.md4.3 KB
- 参考文档references/dashboard-map.md1.7 KB
- 参考文档references/dcf-architecture.md6.2 KB
- 参考文档references/industry-playbooks.md4.2 KB
- 参考文档references/integrity-controls.md4.7 KB
- 参考文档references/model-math.md3.1 KB
- 参考文档references/output-and-review.md4.8 KB
- 参考文档references/output-spec.md6.0 KB
- 参考文档references/p0-integrations.md2.2 KB
- 参考文档references/plan-schema.md5.8 KB
- 参考文档references/qa-checks.md3.2 KB
- 参考文档references/reference-router.md2.2 KB
- 参考文档references/sensitivity-and-scenarios.md4.2 KB
- 参考文档references/valuation-judgment.md5.9 KB
- 参考文档references/wacc-terminal-value.md4.9 KB
- 脚本scripts/build_banker_formula_workbook_runtime48.4 KB
- 脚本scripts/build_banker_formula_workbook.py623 B
- 脚本scripts/plan_validation.py18.4 KB
- 脚本scripts/run_pipeline.py10.6 KB
- 脚本scripts/runtime_loader.py1.1 KB
- 脚本scripts/skill_core.py58.0 KB
- 脚本scripts/validate_plan.py1.4 KB
- 脚本scripts/xlsx_writer.py10.5 KB
03 SKILL.md 正文
DCF Model Builder
Skill Configuration
Common Skill Instructions
MANDATORY: Before searching connectors, retrieving evidence, or drafting output, read and apply the shared runtime contract in ../public-equity-investing/SKILL.md## Cross-Skill Runtime Contract. Then check the router skill map and ../../shared/plugin-routing-playbook.md for adjacent skills that should be sequenced with this workflow. Do not run user-context setup or inspection during ordinary workflow work; route only explicit saved-context, source-setup, onboarding, or automation-setup requests to ../user-context/SKILL.md.
Relevant Dependency Categories
These are the source categories most likely to matter for this workflow. Use the router contract to resolve only the categories the task actually needs, prefer user-named sources first, and state any material source limitation.
Company Filings & IRMarket Data & EstimatesPortfolio Models & Trackers
Deliverable Intake
Before source gathering or analysis for a new substantive hero deliverable, load ../../shared/deliverable-intake-policy.md and use its adaptive request_user_input preflight for materially unresolved format, depth, audience/use, or focus choices. Reuse resolved preferences in downstream steps; when acting only as input to an owning workflow, do not re-prompt.
Own DCF-specific valuation builds: FCFF/FCFE method choice, WACC or cost of equity, terminal value, EV-to-equity bridge, per-share value, valuation sensitivities, QA checks, and a formula-first XLSX workbook.
Hard Contract
- Default model-build artifact level is
banker_formula_workbook; use the formula workbook path for normal public-equity DCF build, refresh, valuation, and model-package requests. - Formula mode is executed through
scripts/build_banker_formula_workbook.pyand must be labeledbanker_formula_workbookonly when its run log, workbook inspection, andmodel_citations.jsonprove the formula builder ran successfully. - Deterministic scripts remain available for controlled computed values, smoke tests, or explicit lightweight support exports. They produce
output/model.xlsx,output/plan.json,output/run_log.json, optional support noteoutput/support_note.md, andoutput/manifest.json. Legacyoutput/report.mdis written only with--write-report-md. - Formula scripts produce
banker_formula_workbook.xlsx,banker_formula_workbook_run_log.json,model_citations.json, andmanifest.jsonin the selected output directory. The workbook is the hero deliverable; citation JSON, run logs, and manifests are support artifacts. - Formula and deterministic workbooks must start with a
Covertab that functions as an investment landing page: issuer, current price versus implied value, valuation range, case outputs, DCF bridge, WACC/terminal assumptions, sensitivity drivers, source posture, warnings/hard failures, and workbook map. run_log.jsonandmanifest.jsonare required reliability artifacts; hard failures forcenot-decision-ready.- Never delete, overwrite, or mutate source data unless explicitly requested. Write new outputs under
output/or a clearly named copy. - Keep material inputs source-labeled:
reported,company_guidance,consensus,management_case,user_provided,connected_app,web_research,analyst_estimate,placeholder, orderived. - If placeholders, stale data, unsupported WACC, missing share count/net debt, or weak analyst estimates drive value, mark the output no higher than
screen-grade.
Routing
Use this skill when the deliverable is a DCF build, refresh, rerun, valuation sensitivity, reverse DCF, or DCF model package. Route away when the primary ask is workbook audit/debug (model-audit-tieout), source hierarchy (financial-source-of-truth), raw financial normalization (financials-normalizer), raw spreadsheet cleanup (excel-data-cleaner), expanded stress architecture (scenario-sensitivity-generator), trading comps (comps-valuation), or memo/deck polish.
For complex tasks, split the work into source-of-truth, accounting normalization, operating forecast, cost of capital, terminal value/ROIC, scenario valuation, and audit workstreams. If sub-agents are unavailable, emulate those workstreams explicitly before running the pipeline.
When To Invoke Support
Load shared/support-layer-routing-contract.md when source/data/QC/style support is needed. Use financial-source-of-truth before relying on load-bearing market data, filings, guidance, consensus, share count, net debt, WACC, or terminal assumptions. Use financials-normalizer before plan creation when historical financials, KPI schedules, segment data, guidance, consensus/provider exports, share count, net debt, or capital allocation inputs are messy. Use excel-data-cleaner before importing malformed tables. Use model-audit-tieout for formula integrity, external links, recalc/cache posture, and final tie-out; use deck-report-qc only for circulation packs. Support artifacts stay secondary to the formula workbook or HTML report.
Input Handling
- Preferred input is a
plan.jsonmatching the schema. - No context: run
assets/plan_template.jsononly as illustrative and label placeholders. - Named company with no data: gather connected/public data when allowed; otherwise ask targeted questions or build a caveated screen-grade plan.
- Partial context: preserve user assumptions, fill only non-economic structure when obvious, and label gaps.
- Full plan: validate first; do not silently change conclusion-driving assumptions.
- When the investment question turns on an operating KPI such as bookings, GBV, take rate, volumes, users, units, or retention, load
references/industry-playbooks.mdand build from the sector-relevant driver where the workbook supports it. If formula mode cannot represent that driver schedule, disclose the proxy prominently on the Cover or Executive Summary and in Source Notes, and keep the model no higher thanscreen-grade; do not silently substitute a generic revenue CAGR.
Required Workflow
- Create or identify a valid
plan.json. - Validate with
python3 scripts/validate_plan.py path/to/plan.json. - Default path: execute
python3 scripts/build_banker_formula_workbook.py path/to/plan.json --output-dir outputand inspectbanker_formula_workbook_run_log.json. - Use
python3 scripts/run_pipeline.py path/to/plan.jsononly for controlled computed values, smoke tests, explicit lightweight support exports, or when formula mode fails and the fallback is clearly labeleddeterministic_export. - Review the relevant run log for hard failures, warnings, checks, source basis, workbook inspection, and P0/model handoff.
- Render and visually inspect the generated workbook before delivery. Confirm that template example data is absent from a named-company model, actual/estimate periods align across tabs, core outputs and sources are legible, and checks disclose any unsupported driver or roll-forward.
- Deliver the workbook as the hero artifact and link the normalized plan, run log, manifest, citation ledger, and optional support note as audit files.
Smoke test:
python3 scripts/validate_plan.py assets/plan_template.json
python3 scripts/build_banker_formula_workbook.py assets/plan_template.json --output-dir /tmp/public-equity-investing-dcf-formula
python3 scripts/run_pipeline.py assets/plan_template.json
Formula Mode Guardrail
Load references/banker-formula-workbook-contract.md before using formula mode. If the formula builder fails, the template is missing, required tabs/formulas/styles are not present, external links appear, or model_citations.json is absent, do not describe any fallback artifact as a banker formula workbook.
Status Labels
Use exactly: decision-grade, senior-review-ready, screen-grade, not-decision-ready, blocked. Use blocked only before a run log exists; after execution, hard failures mean not-decision-ready.
Deferred Reference Router
Load references/reference-router.md only when deeper schema, math, QA, integration, sector, or senior valuation judgment guidance is needed. The router preserves the detailed reference map without loading every reference during invocation.
Equity Valuation PM Standard
Load shared/equity-valuation-pm-standard.md and shared/pm-judgment-heuristics.md for substantial model, valuation, scenario, model-update, or audit work.
The output must state what the current stock price implies, the variant estimate path, whether upside is driven by fundamentals, multiple expansion, mix, capital return, sentiment, or event probability, what breaks first in downside, what changes target, rating, sizing, hedge, trim, exit, or watchlist status, and what evidence is missing.
Keep equity valuation as the center of gravity. Debt is allowed only as an input to common-equity value through net debt, cost of debt, leverage, liquidity, refinancing risk, or downside equity impairment. Use Credit Markets for bond comps, loan comps, CDS, spread/yield relative value, covenant-package analysis, debt-security valuation, recovery waterfall, restructuring valuation, creditworthiness, private-credit / public-credit instrument underwriting, or distressed claim valuation.
04 同仓库更多 Skills
skill-installer
Install Codex skills into $CODEX_HOME/skills from a curated list or a GitHub repo path. Use when a user asks to list installable skills, install a curated skill, or install a skill from another repo (including private repos).
skill-creator
Create or update a Codex skill with appropriately scoped instructions and any needed supporting resources.
review-agent
Perform a read-only, defect-first review of a specified code change and return every actionable finding. Use when another agent delegates review of uncommitted changes, a base-branch diff, a commit, or custom review instructions.
plugin-creator
Create and scaffold plugin directories for Codex with a required `.codex-plugin/plugin.json`, optional plugin folders/files, valid manifest defaults, and personal-marketplace entries by default. Use when Codex needs to create a new personal plugin, add optional plugin structure, generate or update marketplace entries for plugin ordering and availability metadata, or update an existing local plugin during development with the CLI-driven cachebuster and reinstall flow.
openai-docs
Use for Codex models/pricing, scheduled tasks, skills, settings, setup, troubleshooting, customization, automations, and self-knowledge—including 'you,' 'your,' 'this app,' or 'this coding agent' when they refer to Codex—and for OpenAI APIs/products and ChatGPT Work. Also use for model choice/migration, prompting, SDKs, Responses, Realtime, agents, evals, and Chat/Work/Codex comparisons. Do not use for generic app/software tasks that merely mention Codex.
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output should be a bitmap asset rather than repo-native code or vector. Do not use when the task is better handled by editing existing SVG/vector/code-native assets, extending an established icon or logo system, or building the visual directly in HTML/CSS/canvas.