Skip to content

Add Blue Origin Leadership Principles - #87

Merged
tig merged 7 commits into
mainfrom
add-blue-origin
Oct 9, 2026
Merged

tig merged 7 commits into
mainfrom
add-blue-origin

Conversation

@tig

@tig tig commented Oct 9, 2026 •

Copy link
Copy Markdown
Member

Add Blue Origin as a company set, transcribed from the published Leadership Principles on https://www.blueorigin.com/careers/work-culture.

Block 9000. Fourteen principles, in the order the page lists them:

  1. Passion for our Mission
  2. Customer Focus
  3. Deliver Results
  4. Embrace Team Blue
  5. Resourceful
  6. Bias for Action
  7. Ownership
  8. Operational Excellence
  9. Earn the Trust of Others
  10. Hire and Develop the Best
  11. Insist on the Highest Standards
  12. Technical Ambition and Simplicity
  13. Practice Humility
  14. Have Backbone; Disagree and Commit

Definitions are quotations of that page. Passion for our Mission keeps the opening sentence, with the en dash folded to an ASCII hyphen. Practice Humility includes the future-lessons sentence, with the dashes folded to ASCII hyphens. Earn the Trust of Others stops before its last sentence, which uses a dash, so that definition stays a contiguous leading excerpt. Calibration rows are authored. Every principle sits on a facet with generated rows, so Porridge has a table.

Facet maps:

  • Passion for our Mission -> care-about-the-mission
  • Customer Focus -> customer-obsession
  • Deliver Results -> deliver-results
  • Embrace Team Blue -> pull-people-together (new)
  • Resourceful -> frugality
  • Bias for Action -> bias-for-action
  • Ownership -> acts-like-an-owner, safety (new)
  • Operational Excellence -> prevent-the-next-one, dive-deep
  • Earn the Trust of Others -> listens-with-an-open-mind (new)
  • Hire and Develop the Best -> best-person-for-the-role, hire-and-develop-the-best, team-for-the-outcome
  • Insist on the Highest Standards -> highest-standards, safety (new)
  • Technical Ambition and Simplicity -> invent-and-simplify
  • Practice Humility -> stay-humble
  • Have Backbone; Disagree and Commit -> have-backbone-disagree-and-commit

pull-people-together is new because the hiring-bar table and the development table each miss part of pulling people together, healthy tension, and timely critical feedback. Its source refs include feedback that helps and energy from the group, the same behaviors as the generated table.

listens-with-an-open-mind is new because the earn-trust table covers slipped dates, credit, and stakeholder bad news. Blue Origin's text is listening, examining convictions, openness, and diverse viewpoints.

safety is new because Ownership and Insist on the Highest Standards both define the same safety behavior. Ownership says never compromise safety. Insist on the Highest Standards says keep safety front and center. One facet holds both row refs and a generated safety table, so both principles inherit it. The owner rows stay on acts-like-an-owner. The raise-the-bar rows stay on highest-standards.

Operational Excellence maps attention to detail onto the existing dive-deep facet, including the attention-to-detail row. The local prevent-the-next-one slice and the shared facet both include root cause, defects down the line, stay fixed, and mechanisms. Consumers read cross-company facet ids, so Operational Excellence gets the dive-deep detail-checking table and the prevent-the-next-one table, plus the BIQ matches on those facets.

Calculated risk taking, working backwards from the customer, and rise to the occasion are facets of their principles, the same way Amazon models those terms. Two over examples now show high standards and resourcefulness taken too far.

Merge this before the BIQ and Porridge follow-ups. kindelwww pins come later, only to main commits, after this is merged.

…ership Principles on https://www.blueorigin.com/careers/work-culture.

Block 9000. Fourteen principles, in the order the page lists them:

1. Passion for our Mission
2. Customer Focus
3. Deliver Results
4. Embrace Team Blue
5. Resourceful
6. Bias for Action
7. Ownership
8. Operational Excellence
9. Earn the Trust of Others
10. Hire and Develop the Best
11. Insist on the Highest Standards
12. Technical Ambition and Simplicity
13. Practice Humility
14. Have Backbone; Disagree and Commit

Definitions are quotations of that page. Two sentences on the page used dashes, so those sentences were left out of the excerpts (Earn the Trust of Others, Practice Humility). Calibration rows are authored. Every principle sits on a facet with generated rows, so Porridge has a table.

Facet maps:

- Passion for our Mission -> care-about-the-mission
- Customer Focus -> customer-obsession
- Deliver Results -> deliver-results
- Embrace Team Blue -> pull-people-together (new)
- Resourceful -> frugality
- Bias for Action -> bias-for-action
- Ownership -> acts-like-an-owner
- Operational Excellence -> prevent-the-next-one
- Earn the Trust of Others -> earn-trust
- Hire and Develop the Best -> best-person-for-the-role, hire-and-develop-the-best, team-for-the-outcome
- Insist on the Highest Standards -> highest-standards
- Technical Ambition and Simplicity -> invent-and-simplify
- Practice Humility -> stay-humble
- Have Backbone; Disagree and Commit -> have-backbone-disagree-and-commit

pull-people-together is new because the hiring-bar table and the development table each miss part of pulling people together, healthy tension, and timely critical feedback.

Merge this before the BIQ and Porridge follow-ups. kindelwww pins come later, only to main commits, after this is merged.
@tig

tig commented Oct 9, 2026

Copy link
Copy Markdown
Member Author

@codex review

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 9, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-09T15:00:18.922988Z ded77ad Draft marked ready
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@tig
tig requested a balanced review from Copilot October 9, 2026 12:23

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 668ab4b7b0

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread data/blue-origin/practice-humility.json Outdated
Comment thread SCHEMA.md Outdated
Comment thread README.md Outdated

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟡 Changes recommended

Several term classifications violate the schema, and some calibration rows have incorrect under/over polarity.

6 open findings
What changed in this PR

Adds Blue Origin’s 14 Leadership Principles, authored calibration data, and cross-company facet mappings.

Changes:

  • Registers Blue Origin with ID block 9000.
  • Adds principle records and facet mappings, including pull-people-together.
  • Updates generated indexes, documentation, audits, and tests.
File Description
scripts/​companies.py Registers Blue Origin metadata.
SCHEMA.md Documents its block, ordering, and company ID.
README.md Adds Blue Origin to the corpus list.
data/​index.json Indexes all 14 principles.
data/​facets.json Adds mappings and generated calibration.
data/​blue-origin/​bias-for-action.json Adds Bias for Action.
data/​blue-origin/​customer-focus.json Adds Customer Focus.
data/​blue-origin/​deliver-results.json Adds Deliver Results.
data/​blue-origin/​earn-the-trust-of-others.json Adds Earn the Trust of Others.
data/​blue-origin/​embrace-team-blue.json Adds Embrace Team Blue.
data/​blue-origin/​have-backbone-disagree-and-commit.json Adds disagreement and commitment behavior.
data/​blue-origin/​hire-and-develop-the-best.json Adds hiring and development behavior.
data/​blue-origin/​insist-on-the-highest-standards.json Adds standards behavior.
data/​blue-origin/​operational-excellence.json Adds operational excellence behavior.
data/​blue-origin/​ownership.json Adds ownership behavior.
data/​blue-origin/​passion-for-our-mission.json Adds mission behavior.
data/​blue-origin/​practice-humility.json Adds humility behavior.
data/​blue-origin/​resourceful.json Adds resourcefulness behavior.
data/​blue-origin/​technical-ambition-and-simplicity.json Adds technical ambition behavior.
tests/​fixtures/​facet-audit.json Records facet decisions for all principles.
tests/​test_facets.py Extends ownership facet expectations.

🧠 Review effort: Balanced


💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Comment thread data/blue-origin/bias-for-action.json Outdated
Comment thread data/blue-origin/earn-the-trust-of-others.json Outdated
Comment thread data/blue-origin/have-backbone-disagree-and-commit.json
Comment thread data/blue-origin/have-backbone-disagree-and-commit.json Outdated
Comment thread data/blue-origin/hire-and-develop-the-best.json
Comment thread data/blue-origin/operational-excellence.json
@tig

tig commented Oct 9, 2026

Copy link
Copy Markdown
Member Author

@codex review

@tig
tig requested a balanced review from Copilot October 9, 2026 12:56

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 114c143473

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread data/blue-origin/passion-for-our-mission.json Outdated
Comment thread data/facets.json Outdated

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🔵 Needs a closer look

Several term classifications, one facet mapping, and two calibration rows violate repository semantics.

0 open findings

6 resolved since last review
Previously missed (6)

In code that hasn't changed since last review

Medium severity Model calculated risk taking as a bias-for-action facet

data/​blue-origin/​bias-for-action.json:37

calculated risk taking is one slice of Bias for Action, excluding the broader speed and study-delay behavior. Classifying it as equivalent makes consumers return the entire principle, contrary to SCHEMA.md:267-281; the nearly identical Amazon principle models this exact term as a facet at data/amazon/bias-for-action.json:55-62. Convert it to a facet over the applicable risk rows.

Medium severity Model working backwards as a customer-focus facet

data/​blue-origin/​customer-focus.json:42

working backwards from the customer names only one slice of Customer Focus, while the principle also covers earning and keeping trust. Marking it equivalent makes term consumers return every row, contrary to SCHEMA.md:267-281; the same term is correctly modeled as a facet in data/amazon/customer-obsession.json:73-79. Convert this term to a facet over the applicable working-backwards rows.

Medium severity Model rising to the occasion as a resilience facet

data/​blue-origin/​deliver-results.json:42

rise to the occasion describes the resilience-after-setbacks slice, not the key-inputs and customer-commitment portions of this principle. As equivalent, it returns the whole principle in violation of SCHEMA.md:267-281; data/amazon/deliver-results.json:75-81 models this same term as a facet. Convert it to a facet over the relevant resilience rows.

Medium severity Replace unrelated rejection example with a high-standards failure

data/​blue-origin/​insist-on-the-highest-standards.json:81

Rejecting work because someone else produced it is a not-invented-here failure, not high standards taken too far. This leaves the row without an over-indexed calibration anchor; use a case where raising or enforcing the bar beyond useful quality prevents completion.

Medium severity Replace hidden-spend example with an overextended frugality failure

data/​blue-origin/​resourceful.json:67

Hiding extra spend avoids the constraint rather than over-indexing resourcefulness, so this row has two under-indexed failure modes and no valid over anchor. The over case should show frugality or constraint-driven invention taken too far.

Medium severity Model shared listening as a narrower facet

data/​facets.json:947

Adding 9009 here makes consumers inherit every generated earn-trust row, but the retained Blue Origin definition only covers open-minded listening and diverse viewpoints. Rows such as a-date-you-gave-slips, credit-after-a-shared-win, and comparing-the-teams-work are unsupported by that text, so this violates AGENTS.md:102-104 and exposes unrelated calibration and BIQ content. Model the shared listening behavior as a narrower facet with its own generated rows instead.

🧠 Review effort: Balanced

…acet, and calibration.

Passion for our Mission now keeps the opening sentence, with the en dash folded to an ASCII hyphen. Earn the Trust of Others sits on listens-with-an-open-mind instead of the broad earn-trust table. Calculated risk taking, working backwards from the customer, and rise to the occasion are facets. Two over examples now show the behavior taken too far.
@tig

tig commented Oct 9, 2026

Copy link
Copy Markdown
Member Author

@codex review

@tig
tig requested a balanced review from Copilot October 9, 2026 13:29

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 87abea94ec

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread data/blue-origin/deliver-results.json Outdated
Comment thread data/facets.json Outdated
…tion.

The published word is naiveté, in Unicode NFC. ASCII folding is for punctuation, not this accent. The pull-people-together row now puts people onto one direction instead of manufacturing a dependency.
@tig

tig commented Oct 9, 2026

Copy link
Copy Markdown
Member Author

@codex review

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🔵 Needs a closer look

One quoted definition changes the source spelling beyond the schema’s permitted normalization.

0 open findings

Previously missed (1)

In code that hasn't changed since last review

Medium severity Preserve the source spelling and accent in quoted definitions

data/​blue-origin/​deliver-results.json:7

The official source spells this word naiveté, but the transcription drops the accent. SCHEMA.md:247-254 allows ASCII folding only for punctuation, not letters, so this conflicts with the PR's claim that definitions are quotations. Preserve the source spelling.

🧠 Review effort: Balanced

@tig
tig requested a balanced review from Copilot October 9, 2026 13:36
@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. Another round soon, please!

Reviewed commit: 8ad02e3c27

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🔵 Needs a closer look

Safety and attention-to-detail behaviors are omitted from the cross-company facet composition consumed by downstream applications.

0 open findings

Previously missed (2)

In code that hasn't changed since last review

Medium severity Add shared safety facet for Ownership and Highest Standards

tests/​fixtures/​facet-audit.json:567

Safety is modeled here as only a constraint on Ownership, but both Ownership and Insist on the Highest Standards define and calibrate the same safety behavior (safety-is-not-negotiable and safety-front-and-center). Leaving it off the cross-company map means neither principle inherits a safety table, contrary to the shared-behavior composition rule in AGENTS.md:45-49 and SCHEMA.md:154-181. Add one safety facet containing both principles and row refs, generate its table, and project it into the audit and index.

Medium severity Map attention-to-detail facet into Operational Excellence

tests/​fixtures/​facet-audit.json:573

This mapping drops the explicit attention-to-detail slice: the record defines attention-to-detail as a local facet, while the existing dive-deep facet already supplies generated detail-checking rows. Because consumers use only the cross-company facet IDs, Operational Excellence will expose recurrence-prevention examples but no attention-to-detail examples or BIQ matches. Add 9008 and its attention-to-detail row to dive-deep, include that facet here, and rebuild the index.

🧠 Review effort: Balanced

…attention to detail onto dive-deep.

Ownership and Insist on the Highest Standards both calibrate safety, so they share one safety facet with generated rows. Operational Excellence keeps prevent-the-next-one and also sits on dive-deep for its attention-to-detail row.
@tig
tig requested a balanced review from Copilot October 9, 2026 13:54
@tig

tig commented Oct 9, 2026

Copy link
Copy Markdown
Member Author

@codex review

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. Bravo.

Reviewed commit: 97baa89e19

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🔵 Needs a closer look

The prevent-the-next-one mappings omit the explicitly assigned defects-down-the-line behavior.

0 open findings

Previously missed (2)

In code that hasn't changed since last review

Medium severity Add defects-down-the-line to prevent-the-next-one slice

data/​blue-origin/​operational-excellence.json:42

This local prevent the next one slice also omits defects-down-the-line, despite the PR description and facet-audit reason assigning that behavior to this facet. Term consumers will therefore exclude the defect-handoff calibration when users ask for this slice.

Medium severity Include 9008 row in prevent-the-next-one facet

data/​facets.json:2955

defects-down-the-line is omitted even though the PR description and the new facet-audit entry explicitly place defects sent down the line on prevent-the-next-one. As written, this authored behavior is not source material for that cross-company facet; add its 9008 row reference here.

🧠 Review effort: Balanced

The Operational Excellence slice and the shared facet both name the defects-down-the-line row, so the handoff calibration stays with root cause, stay fixed, and mechanisms.
@tig
tig requested a balanced review from Copilot October 9, 2026 14:39
@tig

tig commented Oct 9, 2026

Copy link
Copy Markdown
Member Author

@codex review

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: d0b2528dc6

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread data/facets.json

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🔵 Needs a closer look

The extensive editorial calibration and facet modeling warrants final human validation against the published source.

0 open findings

🧠 Review effort: Balanced

The generated table already covers feedback that helps and energy from the group. Those human rows are now source refs, so a later regeneration keeps them.
@tig
tig requested a balanced review from Copilot October 9, 2026 14:48
@tig

tig commented Oct 9, 2026

Copy link
Copy Markdown
Member Author

@codex review

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. Breezy!

Reviewed commit: ded77ad08c

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟢 Approval recommended

The company records, facet composition, generated index, documentation, and audit coverage are internally consistent.

0 open findings

🧠 Review effort: Balanced

@tig
tig marked this pull request as ready for review October 9, 2026 14:57
@tig tig self-assigned this Oct 9, 2026
@tig
tig merged commit 270795e into main Oct 9, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants