GitHub - bernardohcrocha/persistia-for-claude-code: Give Claude Code the persistent memory it was missing and turn it into your operational co-pilot. · GitHub
/" data-turbo-transient="true" />
Skip to content
Search or jump to...
Search code, repositories, users, issues, pull requests...
-->
Search
Clear
Search syntax tips
Provide feedback
--><br>We read every piece of feedback, and take your input very seriously.
Include my email address so I can be contacted
Cancel
Submit feedback
Saved searches
Use saved searches to filter your results more quickly
-->
Name
Query
To see all available qualifiers, see our documentation.
Cancel
Create saved search
Sign in
/;ref_cta:Sign up;ref_loc:header logged out"}"<br>Sign up
Appearance settings
Resetting focus
You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.
Dismiss alert
{{ message }}
bernardohcrocha
persistia-for-claude-code
Public
Notifications<br>You must be signed in to change notification settings
Fork
Star<br>227
main
BranchesTags
Go to file
CodeOpen more actions menu
Folders and files<br>NameNameLast commit message<br>Last commit date<br>Latest commit
History<br>73 Commits<br>73 Commits
scripts
scripts
templates
templates
.gitignore
.gitignore
ARCHITECTURE.md
ARCHITECTURE.md
FRAMEWORK.html
FRAMEWORK.html
LICENSE
LICENSE
Persistia.png
Persistia.png
README.md
README.md
setup.sh
setup.sh
View all files
Repository files navigation
Persistia
Self-updating persistent memory for Claude Code.
Give Claude Code the persistent memory it was missing and turn it into your operational co-pilot.
Claude Code has no operational memory. Every session starts from zero. You re-explain your stack, paste context, repeat yourself.
Persistia fixes that. On first run, it scans your entire project — code, architecture, docs, and configs. Reads your .env files and uses credentials you already have. Stores everything in small, focused files with a navigation index. Then monitors changes via git diff: 1,000 files, 3 changed, it reads 3. Token-efficient by design.
The result: an agent that always knows your full project and stays up-to-date automatically — even as you ship new features, add docs, or change your stack.
Minimum input. Maximum output.
Install
Open Claude Code in your project root folder , then run:
curl -fsSL https://raw.githubusercontent.com/bernardohcrocha/persistia-for-claude-code/main/setup.sh | bash
Claude scans your project and asks only what it can't find. No forms. No config files. Just a conversation.
Requires: git · Node.js 18+
Optional: GitHub CLI (gh) — automatically creates a private brain repository with cloud backup on every update. Format your machine, git clone the brain repo, continue exactly where you left off.
Scheduler included: installs automatically (launchd on macOS, systemd on Linux).
How it works
Initial indexing — on first run, scans your entire project: code, docs, configs, and environment files. Asks only what it can't find on its own. Uses credentials you already have — no manual reconnection.
Self-updating memory — runs git diff daily. Only what changed gets read. No unnecessary context loaded.
Permanent skills — say it once → written to _brain/skills/, loaded at every future session
Autonomous tasks — schedule any task in plain language, runs automatically with full context. Each task in _brain/tasks/queue.json supports batch_size (cap items per run — prevents timeouts on large lists) and timeout_minutes (default: 5). Easy to customize.
Proactive scan — every 3 days when idle, it scans metrics, customers, and channels and flags what it notices. Always suggests, never acts unilaterally.
Live dashboard — open _brain/dashboard.html, auto-refreshes every 5 min
Shareable memory — its own private git repo. Format your machine, git clone the brain repo, continue exactly where you left off. Shareable with your team.
Unlike Hermes Agent, Agent Zero, or OpenClaw — context updates itself. No manual maintenance required.
Everything in _brain/ is plain text — readable, editable, and version-controlled.
Make Claude Code your operational co-pilot
A dashboard shows numbers. This answers questions and executes orders — cross-tools.
Ask — revenue, churn, at-risk accounts, support patterns — across all your tools at once.
Give orders — outreach, reports, follow-ups. Executed using credentials you already have.
Schedule — any task, in plain language, runs automatically with full context.
→ "Which customers dropped usage 30%+ this month? Cross-check their support history and draft a personalized re-engagement message for each."
→ "Every Monday: pull last week's numbers from Stripe, flag anomalies, and queue a follow-up for any account that dropped below quota."
→ "Find signups from the last 30 days with no activity. Visit each...