docs: redraw hero banner + reframe gittensor around the miner workflow#335
Conversation
lead the readme with session auto-capture — the feature that stops each new agent session from re-doing the last one's work. - banner.svg: add a "sessions capture themselves" on-ramp that funnels into the review gate; swap the approve/reject text labels for icons; refresh the example date and kb panel; add a self-capturing pill. - gittensor.md + readme gittensor section: reframe from "cited record of the subnet's scoring rules" to a real miner workflow — capture what a session learns about a target repo, approve it, recall it next session so the investigation happens once. grounded in the real sn74 loop (merged prs, repo allocation, a shuffling whitelist). - demo.gif/tape + example-session.md: the auto-capture walkthrough. - .claude/settings.json + commands: reference capture hooks and the four vouch slash commands. settings.local.json intentionally omitted.
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
There was a problem hiding this comment.
Pull request overview
This PR is a documentation and repository-config change that reframes vouch's messaging around automatic Claude Code session capture and repositions the Gittensor (SN74) template around a concrete "miner workflow" (capture what a session learns about a target repo → approve it → recall it next session). It also adds vouch's own repo-local Claude Code config (.claude/settings.json + four slash commands) so the project dogfoods the capture/recall hooks. No source or protocol behavior changes; the underlying capture/recall/CLI surface already exists in the codebase.
Changes:
- Rewrites the README intro, "Why/When", Gittensor section, CLI reference, and Claude Code wiring to lead with mechanical (no-LLM), review-gated session auto-capture and
vouch recall. - Replaces the manual propose→approve walkthrough (
example-session.md,demo.tape) and redrawsbanner.svgaround the capture → review → commit → recall loop; rewritesdocs/gittensor.mdas a miner re-investigation loop. - Adds root
.claude/settings.json(SessionStart/PostToolUse/SessionEnd hooks + read-only kb_* allowlist) and four.claude/commands/*.mdslash commands.
Reviewed changes
Copilot reviewed 10 out of 12 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
| README.md | Reframes pitch around auto-capture + recall; adds install-mcp, capture, recall to CLI docs and comparison table |
| docs/gittensor.md | Rewrites from "SN74 decision-memory" to a miner re-investigation/capture loop; adds gitt client vs vouch layering |
| docs/example-session.md | Replaces manual example with the capture → finalize → approve → recall walkthrough |
| docs/demo.tape | Re-scripts the VHS demo to the auto-capture loop |
| docs/banner.svg | Redraws hero banner with an auto-capture on-ramp, icon approve/reject, refreshed KB panel |
| docs/tutorials/README.md | Updates the example-session link wording |
| .claude/settings.json | New repo-local hooks + kb_* read allowlist (SessionStart omits finalize-all/recall vs adapter) |
| .claude/commands/*.md | Adds status/recall/resolve-issue/propose-from-pr slash commands (identical to adapter copies) |
Notes on the two comments: the example-session.md claim that "credentials … never get buffered" is overstated because Bash observations store the command string verbatim (up to 200 chars); and the root .claude/settings.json SessionStart diverges from the shipped adapter (no finalize-all fallback or recall), which has operational consequences for this repo's own dogfooding. Several external factual claims (SN74 scoring specifics, the gitt client commands, the entrius/gittensor repo URL) could not be verified from the codebase and warrant human confirmation.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| Observations are one-line summaries and file names — **not** full tool | ||
| output — so credentials and large blobs never get buffered. |
| "hooks": [ | ||
| { | ||
| "type": "command", | ||
| "command": "vouch status --json || true" | ||
| }, | ||
| { | ||
| "type": "command", | ||
| "command": "vouch capture banner || true" | ||
| } | ||
| ] |
lead the readme with session auto-capture — the feature that stops each new agent session from re-doing the last one's work.
What changed
Why
What might break
VEP
Tests
make checkpasses locally (lint + mypy + pytest)CHANGELOG.mdupdated under## [Unreleased]