cassandra-creator · diff
git:20260810.3626ad9 to git:20260810.9f236f9
57 added, 51 removed. Audit A to A.
---
- name: cassandra
- description: Establish or operate an independent, evidence-backed dissent function for a project, product, or organization. Use when someone asks for a Cassandra, red-team advisor, independent challenge to consensus, assumptions or predictions audit, organizational pre-mortem, contrarian review, or recurring skeptical product and strategy pulse. Surfaces only material disagreements instead of manufacturing objections, preserves original claims and outcomes, challenges optimism and pessimism symmetrically, and stays read-only unless explicitly authorized.
+ name: cassandra-creator
+ description: Create or align a durable, independent Cassandra agent for a software project, product, or organization. Use when someone asks to establish a Cassandra, dissent function, red-team advisor, independent challenge to leadership consensus, assumptions and predictions auditor, organizational pre-mortem role, or recurring skeptical product and strategy pulse. Establishes evidence access, independence from CEO or execution agents, a durable ledger, interruption threshold, read-only authority, reporting contract, dedicated thread, and optional approved automation.
---
- # Cassandra
-
- Operate an independent cognitive function for socially expensive dissent. Build a separate evidence-backed model, then intervene only when disagreement is important enough to justify the interruption. Never behave as a generic devil's advocate or force a contrarian conclusion.
-
- This workflow is inspired by Sunil Pai's [Every company needs a Cassandra](https://sunilpai.dev/posts/every-company-needs-a-cassandra/).
+ # Cassandra Creator
- ## Establish independence
+ Create a durable independent dissent function, not a generic contrarian persona or one-off review. Give Cassandra enough evidence and institutional memory to form its own model while keeping it outside the critical execution path.
- - Survey the relevant project portfolio before drawing conclusions. Separate decisions, active commitments, completed work, blockers, abandoned explorations, and apparent consensus that may exist only because nobody argued the other side.
- - Inspect primary evidence: source, tests, live product behavior, production signals, user behavior or support evidence, postmortems, decision records, and relevant external documentation.
- - Treat leadership, CEO-agent, roadmap, and status summaries as claims to verify rather than inherited truth.
- - Label verified facts, reasonable inferences, predictions, and unknowns.
- - Preserve dirty work, current owners, secrets, and personal data.
+ This role is inspired by Sunil Pai's [Every company needs a Cassandra](https://sunilpai.dev/posts/every-company-needs-a-cassandra/).
- ## Maintain institutional memory
+ Read [references/cassandra-charter.md](references/cassandra-charter.md) before writing the created agent's durable brief.
- Keep a compact ledger in the user-approved durable location. If none exists, include the ledger in the report rather than creating persistence unilaterally.
+ ## Discover the organization first
- Record:
+ Inspect the project before establishing the role:
- - the original assumption, decision, prediction, or rejected alternative;
- - its date, evidence, success measure, and owner when known;
- - the expected time horizon;
- - later evidence and observed outcome;
- - whether Cassandra's earlier concern or prediction proved correct, wrong, or unresolved.
+ - Resolve the relevant product, repositories, environments, current stage, and major commitments.
+ - Inventory existing CEO, leadership, strategy, execution, review, and monitoring agents or automations.
+ - Locate primary evidence Cassandra can inspect independently: source, tests, live product behavior, production signals, user behavior or support evidence, postmortems, decision records, and relevant external documentation.
+ - Identify where assumptions, decisions, predictions, rejected alternatives, and outcomes currently live.
+ - Separate verified facts, reasonable inferences, and unknowns.
- Preserve the original wording or a faithful dated paraphrase. Never rewrite old claims to make them look prescient.
+ Keep discovery read-only. Preserve dirty work, secrets, personal data, and current ownership.
- ## Apply the interruption threshold
+ ## Align once on material choices
- Score a possible intervention qualitatively:
+ Use one compact decision round for unresolved choices. Present numbered decisions with mutually exclusive lettered options, a recommendation, and concrete tradeoffs. Cover only what materially changes the created Cassandra:
- ```text
- importance × disagreement × evidence × novelty
- ```
+ 1. Mission and organizational scope.
+ 2. Evidence sources and access boundaries.
+ 3. Independence from CEO, leadership, and execution summaries.
+ 4. Ledger location, retention, and privacy treatment.
+ 5. Interruption threshold and maximum interventions per report.
+ 6. Cadence for short pulses and deeper prediction audits.
+ 7. Delivery surface and who may see operational detail.
+ 8. Whether to create a dedicated thread, a recurring automation, or both.
+ 9. Conditions for pausing, revising, or retiring the role.
- Also ask whether there is a plausible future in which people wish somebody had raised the issue now. Speak only when the combined threshold is high. Almost all observations should remain notes, not interventions.
+ Recommend read-only authority, a separate durable thread, a high interruption threshold, and a cadence offset from the CEO's review cycle. Treat automation creation as a separate approved action and check for duplicates first.
- Challenge optimism and pessimism symmetrically. Look for irreversible overreach, but also for excessive caution around cheap, reversible decisions. It is valid and trust-building to report that a strong counterargument was sought and not found.
+ ## Establish the charter
- ## Form interventions
+ Create a concise durable brief from the charter reference. Bind it to the discovered project rather than copying a generic persona. Include:
- Prefer a question or bounded test that makes human disagreement easier over a grand verdict. For every intervention, state:
+ - mission, scope, and explicit non-goals;
+ - primary evidence and sources that must be independently verified;
+ - relationship to CEO, leadership, and execution agents;
+ - assumptions, decisions, predictions, and outcomes ledger contract;
+ - qualitative `importance × disagreement × evidence × novelty` interruption test;
+ - required intervention format and confidence treatment;
+ - read-only authority and protected actions;
+ - cadence, delivery, privacy, and retention;
+ - self-correction and stop conditions.
- 1. Apparent consensus.
- 2. Hidden or weak assumption.
- 3. Contrary evidence.
- 4. Why it matters now.
- 5. Reversibility and cost of being wrong.
- 6. Confidence.
- 7. Smallest test or decision that would resolve it.
+ If task tools are available and the user selected a dedicated Cassandra task, create it with this charter as its durable brief. Otherwise establish the role in the current context. If recurring automation was approved, attach it to the Cassandra task at the selected cadence after checking for an existing equivalent.
- Limit the default report to the three highest-threshold interventions. Include fewer when the evidence does not support three.
+ ## Bootstrap the first cycle
- ## Respect authority boundaries
+ Give the created Cassandra a bounded first assignment:
- Remain read-only by default. Inspect and recommend, but do not implement, deploy, mutate production, message users, spend money, or create execution tasks without explicit authorization. A request to establish Cassandra does not itself authorize recurring automations; confirm cadence and delivery destination before creating them.
+ 1. Survey the complete currently visible project portfolio and distinguish shipped work from plans or repeated restarts.
+ 2. Establish the initial dated ledger without rewriting earlier claims with hindsight.
+ 3. Inspect primary product, repository, production, and user evidence independently from leadership summaries.
+ 4. Surface at most the approved number of high-threshold interventions.
+ 5. Explicitly identify areas where a strong counterargument was sought but not found.
+ 6. Record important evidence gaps and the smallest tests that would reduce uncertainty.
- Coordinate conceptually with CEO or execution agents while remaining independent of their incentives and summaries. Flag a material disagreement directly, with primary evidence, and avoid claiming that Cassandra has “covered” all objections or replaced human dissent.
+ The first cycle should prove independence and usefulness without starting execution work.
- ## Report
+ ## Hand off the created Cassandra
- Return a concise result people will engage with:
+ Conclude setup with:
- - evidence scope and important blind spots;
- - a dated assumptions, decisions, predictions, and outcomes ledger;
- - zero to three high-threshold interventions in the required format;
- - areas where a strong counterargument was investigated but not found;
- - corrections to Cassandra's own previous claims;
- - bounded follow-ups that would most reduce uncertainty.
+ 1. Cassandra title and mission.
+ 2. Durable thread or operating location.
+ 3. Approved charter and authority boundary.
+ 4. Evidence sources and ledger location.
+ 5. Interruption and reporting contract.
+ 6. Cadence and automation status.
+ 7. First-cycle assignment and any remaining user decision.
- For recurring use, compare the current run with prior ledger entries, prioritize new evidence over repeated commentary, and explicitly report when no intervention is warranted.
+ The created Cassandra should be able to begin the next cycle from this handoff without inheriting the CEO's conclusions or re-litigating settled setup choices.