Conversation
There was a problem hiding this comment.
Pull request overview
Adds a backlog entry capturing a proposed on-demand skill to download ChatGPT conversation exports via an authenticated in-browser fetch flow, intended to be reusable and not run on any cadence.
Changes:
- Adds a new backlog item describing the “ChatGPT conversation-download skill” scope, outputs, and constraints (on-demand; writes to
drop/; no auto-absorb). - Records the operational details needed to reproduce the download mechanism (required headers, endpoint shape, Playwright usage).
Aaron Otto-108 verbatim: "if it's easy to do we might as well add a chatgpt conversation/chat download skill on demand no cadence backlog" Technique proven Otto-107 (24 MB / 3992-msg / 8-month Amara conversation in one fetch): navigate, extract Bearer JWT from /api/auth/session, call backend-api/conversation/<UUID> with account-id + project-id headers, unwrap double-JSON from browser_evaluate. Packaging as reusable skill lets future invocations skip the auth-header archaeology; Otto-107 just did this work by hand. Scope: extract-conversation-id -> navigate -> fetch-with-headers -> unwrap -> save to drop/ -> emit stats. NOT absorb (separate §33- discipline skill). NOT cadenced (on-demand per Aaron directive). SPOF noted per Otto-106: account/project IDs are workspace-scoped; skill extracts fresh per run rather than hardcoding. Priority P3 convenience; effort S. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
AceHack
force-pushed
the
backlog/chatgpt-conversation-download-skill
branch
from
55dce45 to
226b32f
Compare
This was referenced Apr 26, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Aaron Otto-108: "if it's easy to do we might as well add a chatgpt conversation/chat download skill on demand no cadence backlog"
Mechanism is proven — Otto-107 pulled the full 24 MB / 3992-message / 8-month Amara conversation in one
fetchcall. Packaging as a reusable skill so future Otto instances (or Aaron) can invoke on-demand without redoing the auth-header archaeology.Skill scope (Phase 1 when authored)
fetchinside page with auth headers from the active sessionbrowser_evaluatedouble-JSON encodingdrop/chatgpt-conversations/<ID>.json(drop/ gitignored per PR fix: gitignore drop/ (protection corrects prior memory claim) #299)Does NOT auto-absorb into docs/ (absorb is separate §33-discipline skill). Does NOT trigger on cadence (on-demand per Aaron directive).
SPOF note (per Otto-106)
chatgpt-account-id+chatgpt-project-idare workspace-scoped and may rotate. Skill extracts both fresh each run rather than hardcoding.Priority / effort
P3 convenience; S effort (mechanism proven; packaging +
skill-creatorworkflow authoring is the remaining work).🤖 Generated with Claude Code