CATCH #797
DETECTED AUG 18, 2026 17:14 UTC
THE CATCHscore 9/10
EXHIBIT 797 · FLASH
Aug 18, 2026 17:14 UTC
web app
https://platform.claude.com
Anthropic's Claude console has new frontend strings hinting at a 'claude-3-opus' model ID, internal reasoning token breakdown, and new API endpoints related to sessions, skills, and organizational management.
READY TO POST:
anthropic's frontend code for claude-3-opus is showing up! looks like there's also new debug info on internal reasoning tokens.
seeing new anthropic api endpoints pop up in the claude console, especially around sessions, skills, and some 'beta' features. keeping an eye out!
EVIDENCE — the receipts
⚠️ FROM WEB-APP BUNDLE — may include vendored-dependency strings, verify before posting.
Anthropic Console (https://platform.claude.com) deployed new frontend code [aca8e3c38b49193a -> bdce4b8d634b7e00]: +8384 strings, -3520 strings, 75 interesting
[model_id] (1)
+ claude-3-opus
[feature_flag] (18)
+ /api/organizations/${y?.uuid}/internal_tier
+ /v1/sessions/${a}/resources/${e}?beta=true
+ /v1/sessions/${a}/threads/${e}/archive?beta=true
+ /v1/sessions/${a}/threads/${e}/events?beta=true
+ /v1/sessions/${a}/threads/${e}/stream?beta=true
+ /v1/skills/${a}/versions/${e}/content?beta=true
+ /v1/tunnels/${a}/certificates/${e}/archive?beta=true
+ /v1/tunnels/${a}/certificates/${e}?beta=true
+ /v1/vaults/${a}/credentials/${e}/archive?beta=true
+ /v1/vaults/${a}/credentials/${e}/mcp_oauth_validate?beta=true
+ /v1/vaults/${a}/credentials/${e}?beta=true
+ Breakdown of output tokens by category.
`output_tokens` remains the inclusive, authoritative total used for billing.
This object provides a read-only decomposition for observability — for example,
how many of the billed output tokens were spent on internal reasoning that may
have been summarized be
+ CONSOLE_LOGIN_HCAPTCHA_FRONTEND_ENABLED
+ ExperimentalFeature
+ INTERNAL_getSegmentSpan
+ InternalBackdrop
+ InternalServerError
+ Number of output tokens the model generated as internal reasoning, including
the thinking-block delimiter tokens.
Reflects the raw reasoning the model produced, not the (possibly shorter)
summarized thinking text returned in the response body. Computed by
re-tokenizing the raw reasoning text, so it
[endpoint] (67)
+ /api/organizations/${y?.uuid}/internal_tier
+ /v1/sessions/${a}/resources/${e}?beta=true
+ /v1/sessions/${a}/threads/${e}/archive?beta=true
+ /v1/sessions/${a}/threads/${e}/events?beta=true
+ /v1/sessions/${a}/threads/${e}/stream?beta=true
+ /v1/skills/${a}/versions/${e}/content?beta=true
+ /v1/tunnels/${a}/certificates/${e}/archiv
caught by the 20-minute sweep ·
the live wire · all receipts