Due-Diligence Brief
Produce a structured due-diligence brief on an entity (company, vendor, investment, partner): a planner locks the diligence dimensions and red-flag criteria, a researcher gathers verifiable facts and risk signals per dimension with sources, and a copywriter writes a decision-oriented brief with findings, risks, and an overall read. The ordering matters because diligence without locked dimensions misses a category (e.g. legal, financial, reputational) and a brief without a red-flag rubric buries the risks; fixing the dimensions and what counts as a flag first makes the brief decision-ready and defensible. Use this for vendor/partner due diligence, investment diligence, background briefs on a company or person, risk assessments, and 'should we work with / invest in X' questions.
Steps
Entry step: scope. Each step names the specialist role it wants; the full working prompt is expandable.
- Scope the diligenceplannerentry
lock diligence dimensions + red-flag criteria
Show working prompt
1. Name the entity and the decision the brief informs (invest, partner, buy, hire) in one sentence. 2. Lock the diligence dimensions to cover (e.g. business model & traction, financials/funding, leadership, legal/regulatory, reputation/litigation, security/compliance as relevant) and what each must establish. 3. Define the red-flag rubric: what observations count as a red, amber, or green signal in each dimension. 4. Write an acceptance-criteria checklist (e.g. 'every dimension has a findings section', 'each material fact cites a source', 'risks are rated red/amber/green per the rubric', 'an overall recommendation/read is stated', 'unknowns and information gaps are listed'). 5. Record decisions. Call write_task_note with the dimensions, red-flag rubric, and checklist; write the same to notes/scope.md.
- Gather facts & signalsresearcher
verifiable facts + risk signals per dimension, sourced
Show working prompt
1. For each locked dimension, gather verifiable facts into notes/findings.md: the fact, its source [n] (official filing, registry, reputable reporting), and the date. 2. Apply the red-flag rubric as you go: tag each notable signal red/amber/green with the reason. 3. Prefer primary records (registrations, filings, official statements) over rumor; corroborate any serious negative with a second independent source before flagging it red. 4. Explicitly list what you could NOT find — open questions and information gaps are a diligence output, not a failure. 5. Do not assert unverified rumors as fact; mark them UNVERIFIED. Good looks like a per-dimension findings file with rated signals. Do not write the brief yet.
- Write the briefcopywriter
decision-oriented brief with findings, risks, and a read
Show working prompt
1. Write brief.md: a 1-paragraph bottom-line read up top (overall recommendation + the single biggest risk), then a section per diligence dimension with its findings and rated signals, then a consolidated 'Risks (red/amber/green)' section, then 'Open questions & information gaps'. 2. Every material fact carries its citation [n] and date; never state a serious negative without a corroborating source. 3. Keep it decision-oriented and scannable — a busy reader should get the verdict and risks in 60 seconds. 4. Add a References list from notes/findings.md so every [n] resolves. 5. On a loop-back, fix only the named gaps. 6. Call write_task_note with the brief path and which criteria now pass.
- Evaluatereviewer
Grade the deliverable against every acceptance criterion. All pass → finish; any fail → loop back and fix the gap.
Show working prompt
Open brief.md and check EACH criterion PASS/FAIL: (1) every locked dimension has a findings section; (2) each material fact cites a dated source (spot-check 3 against notes/findings.md); (3) risks are rated red/amber/green per the rubric and serious negatives are corroborated by a second source; (4) an explicit bottom-line read/recommendation is stated up top; (5) an open-questions/information-gaps section lists what could not be confirmed; (6) no UNVERIFIED rumor is presented as fact. A brief that buries the risks, omits a dimension, or states an uncorroborated negative as fact FAILS. Then route — this is the whole point of the loop: - **Every criterion PASSES →** call `advance_task_step({ ref, stepId: "evaluate", next: "finish" })`. - **Any criterion FAILS →** write the specific gaps to notes, then call `advance_task_step({ ref, stepId: "evaluate", next: "write" })` to loop back. The builder fixes exactly those gaps. Never route to `finish` while any criterion is unmet. The build phase's completion gate already blocked a grossly-incomplete deliverable; your job is the judgment an automated check cannot make (does it actually work, read well, look right). After ~3 unproductive loops, stop and report DONE_WITH_CONCERNS so the user can step in. - Finishdeveloper
All acceptance criteria met. Stamp a short summary and report DONE.
Show working prompt
Every acceptance criterion passed. Write a one-paragraph DONE summary to task notes via `write_task_note`: what was built, the deliverable path(s), and a one-line confirmation that each criterion is met. Then report DONE.
Triggers
Phrases that suggest this craftbook to a crew.
- due diligence brief
- diligence on this company
- background check on a vendor
- should we partner with
- risk assessment of an entity
Source
View this craftbook on GitHub · MIT license