Skip to content

docs: use hydra-db for clone and marketplace add? - #12

Open
sudhishramesh wants to merge 1 commit into
hydra-db:mainfrom
sudhishramesh:docs/canonical-github-org
Open

sudhishramesh wants to merge 1 commit into
hydra-db:mainfrom
sudhishramesh:docs/canonical-github-org

Conversation

@sudhishramesh

Copy link
Copy Markdown

Description

Proposal: README / CONTRIBUTING / docs/usage.md still say usecortex/hydradb-claude-code for git clone and /plugin marketplace add. GitHub redirected me to hydra-db/hydradb-claude-code.

I did not touch tenant vs database vocabulary — that looks already in flight.

Related Issue

Closes #

Type of Change

  • Documentation
  • Bug fix
  • New feature
  • Refactoring
  • CI/Build

Checklist

  • I have read the CONTRIBUTING guidelines
  • Linked to an issue (happy to file one if you want)
  • Commits are signed off (git commit -s)
  • npm run check — docs-only, no JS change
  • No secrets

If the marketplace identifier is still supposed to be usecortex/... for existing installs, please say so and I will close this.

Made with Cursor

GitHub already redirects the old usecortex paths. Updating clone and
/plugin marketplace examples so they match the current org.

Signed-off-by: sudhishkr <5901104+sudhishkr@users.noreply.github.com>
@greptile-apps

greptile-apps Bot commented Sep 10, 2026

Copy link
Copy Markdown

Greptile Summary

This documentation-only PR updates clone, upstream, marketplace, issue-label, and Discussions links to the current hydra-db GitHub organization.

  • Aligns installation and contribution instructions with the repository’s current owner.
  • Leaves several machine-readable repository and contribution URLs on the former usecortex owner.
  • Greptile automatically discovered a related ticket that helped explain the purpose of this PR: migrate usecortex-related assets to HydraDB.

Confidence Score: 4/5

The documentation changes are non-blocking and appear safe to merge, though the remaining canonical repository metadata should also be migrated for consistency.

The new URLs match the checked-out repository’s canonical organization, but the partial migration leaves package, plugin, marketplace, and issue-template metadata pointing to the former owner.

Files Needing Attention: README.md

Important Files Changed

Filename Overview
README.md Updates clone and marketplace commands to the current organization, but exposes the broader incomplete metadata migration.
CONTRIBUTING.md Updates upstream, issue-label, and Discussions links while related contribution configuration retains the former organization.
docs/usage.md Consistently updates the public marketplace-add command to the current repository slug.

Reviews (1): Last reviewed commit: "docs: use hydra-db GitHub URLs for clone..." | Re-trigger Greptile

Comment thread README.md

```
/plugin marketplace add usecortex/hydradb-claude-code
/plugin marketplace add hydra-db/hydradb-claude-code

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

P2 Repository URLs Remain Inconsistent

This migration updates the documented marketplace command and related links to hydra-db, but package.json, the plugin and marketplace manifests, and the issue-template configuration still point to usecortex. Consumers and contributors therefore receive conflicting canonical repository and support URLs, and the remaining metadata links will become stale if the legacy redirects are removed. Please update those repository metadata and contribution settings as part of this migration. Greptile automatically discovered a related ticket stating that usecortex-related assets should migrate to HydraDB, which informed this comment.

Source Used: Linear — Migrate usecortex related acc to hydradb

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

@openhack-agent

Copy link
Copy Markdown

🔍 Security Scan Results

⚠️ 1 potential vulnerability detected

View full results on OpenHack

  • 🟢 Low: 1

Findings

🟢 Plugin persists conversation transcripts and recall payloads in project cwd without git-ignore protection outside the plugin repo

Severity: LOW

The HydraDB Claude Code plugin persists conversation state into the project working directory by default. When CLAUDE_PLUGIN_DATA is unset (the normal case), scripts/plugin.mjs fallbackDataDir() resolves the data dir to path.join(cwd, ".hydradb-plugin-data") (line 70), and the wired Stop/UserPromptSubmit hooks run in the user's project cwd. handleStop stores up to 40 turns per session (appendTurn, limit=40) containing the regex-redacted user prompt and the full assistant reply; handleUserPromptSubmit additionally stores the full last-recall additionalContext payload (plus complete memory/knowledge chunks when debug=true) — all written by writeState() to /.hydradb-plugin-data/state.json. The only guard against committing this file is the plugin repo's own .gitignore:1 ('.hydradb-plugin-data/'); ensureDataDir() is a bare mkdir with no self-ignoring .gitignore, no setup step adds the rule to user projects, grep finds zero 'gitignore' references in .mjs, and docs/security.md never mentions commit exposure. In every user repo other than the plugin's own, the transcript store is therefore untracked and unignored — one 'git add -A' / 'git add .' and push away from being committed to a shared or public remote, with only heuristic regex redaction (sanitize.mjs, which misses common secret shapes such as sk_live_, xox*, and connection strings) standing between conversation content and repository history.

Location: scripts/plugin.mjs:526

View details on OpenHack →

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.

1 participant