Social Card Comparison: Default Effort vs. Max Effort
On 2026-03-26, Claude Opus 4.6 generated social card metadata for all ~245
pages in the matheology section — twice, at two different quality levels:
OO (Operates Oddly) — default effort. The agent read the opening of
each page and crafted metadata quickly.
PP (PathProbing) — max effort. The agent read each page in full,
understood its core argument, and crafted metadata with strict character
counting and quality control.
The purpose of this comparison is to help a human reviewer decide: does the
extra effort produce meaningfully better social cards? The answer may differ
by page type — public-facing theological arguments may benefit more than
internal audit logs.
How to Read These Tables
Pick the detail level that matches your interest:
Beginner — For anyone curious about the project. Shows the page name, the
old title vs. the new title, and the old description vs. the new description.
Nothing technical. Just: which one sounds better to you?
Producer — For content managers and reviewers. Adds keyword differences,
character counts, and highlights where the biggest changes occurred. Useful for
deciding which pages need manual attention.
Expert — For the technically curious. All four metadata fields side by side,
exact character counts, quality flags, and review checklists. The raw data for
anyone who wants to audit the process.
Flags — Pages that need human attention. Short titles, long descriptions,
language rule edge cases, or pages where OO and PP diverged significantly.
What the Quality Designations Mean
These designations come from the StayC maturity scale used across this project:
Social Card Comparison: Default Effort vs. Max Effort#
On 2026-03-26, Claude Opus 4.6 generated social card metadata for all ~245 pages in the matheology section — twice, at two different quality levels:
OO (Operates Oddly) — default effort. The agent read the opening of each page and crafted metadata quickly.
PP (PathProbing) — max effort. The agent read each page in full, understood its core argument, and crafted metadata with strict character counting and quality control.
The purpose of this comparison is to help a human reviewer decide: does the extra effort produce meaningfully better social cards? The answer may differ by page type — public-facing theological arguments may benefit more than internal audit logs.
How to Read These Tables#
Pick the detail level that matches your interest:
Beginner — For anyone curious about the project. Shows the page name, the old title vs. the new title, and the old description vs. the new description. Nothing technical. Just: which one sounds better to you?
Producer — For content managers and reviewers. Adds keyword differences, character counts, and highlights where the biggest changes occurred. Useful for deciding which pages need manual attention.
Expert — For the technically curious. All four metadata fields side by side, exact character counts, quality flags, and review checklists. The raw data for anyone who wants to audit the process.
Flags — Pages that need human attention. Short titles, long descriptions, language rule edge cases, or pages where OO and PP diverged significantly.
What the Quality Designations Mean#
These designations come from the StayC maturity scale used across this project:
Code
Name
Meaning for Social Cards
OO
Operates Oddly
Agent read partial page, metadata is functional but may miss the core argument or truncate at awkward boundaries.
PP
PathProbing
Agent read full page, metadata reflects deep understanding, character counts strictly enforced, descriptions end at natural boundaries.