| What it is | A company brain for a team's coding agents. It learns the rules your team decides and serves the ones that apply to each session. | An agent memory system that extracts facts from conversations, consolidates them into observations and recalls them, plus a plugin that adds memory to coding agents. |
|---|
| Where the knowledge comes from | Slack, GitHub, GitLab, Notion, Confluence, Jira, Linear, Google Drive and Granola, plus what agent sessions record. | What agents and users say, extracted into facts. Its coding plugin also reads the repo's git history (commit messages by default, diffs if you ask) and keeps the session transcripts. |
|---|
| Who decides what gets in | You set the gate. Manual holds everything for a person, Hybrid adds routine rules and holds the important ones, Auto adds everything else. A contradiction always waits for a person. | Extraction is automatic, and the coding plugin documents no review step. On Hindsight Cloud, owners and admins manage banks and keys, and members read and use them. |
|---|
| How it reaches the agent | Hooks add the rules that apply to the session before the agent writes a line. The same rules are on an MCP server, the CLI and /harbor in Slack. | Its coding plugin installs hooks that recall on a session's first prompt and retain the transcript when the session stops. Agents can also search through its MCP server, SDKs or HTTP. |
|---|
| When a decision changes | When a push deletes the file a rule is about, the rule is retired, or flagged if other repos still use it. A newer decision that contradicts a rule waits for a person. | Related facts are consolidated into observations, which are refined rather than overwritten when new evidence arrives, and keep the evidence behind them. |
|---|
| Seeing what was used | Every rule shows how often it was served and how often an answer cited it. | Its coding plugin logs, per turn and on your machine, which Hindsight tools the agent called and whether the reply credited Hindsight memory. |
|---|
| Stopping a risky command | Block rules refuse a command before it runs, in Claude Code, Cursor, Codex and GitHub Copilot, and hand the agent something to run instead. | Not part of the product. |
|---|
| Agents | Claude Code, Cursor, Codex and GitHub Copilot (the CLI and VS Code). | 22 coding agents, including Claude Code, Codex CLI, opencode, Cursor CLI and GitHub Copilot CLI, and 30 frameworks and SDKs. |
|---|
| Open source and hosting | The CLI, harborloop, is MIT licensed on npm. The service is hosted. | Open source under MIT and self-hosted with one Docker command. Hindsight Cloud is hosted, with private cloud, your own cloud account or on-prem on Enterprise. |
|---|
| Price | Setup is self-serve and takes two minutes, with no call required. Free costs nothing for your whole team on one source, Pro is $24 per person a month billed yearly or $30 monthly, and Premium is $36, or $45 monthly. | Self-hosted is free. Cloud is billed on tokens with no seat price: $10 per million retained, $0.75 per million recalled and $0.05 a reflect, plus storage after 30 days. Enterprise is custom. |
|---|
| Best for | Engineering teams who want every coding session to start with what the team decided. | Developers who want open-source memory their agents fill and recall on their own. |
|---|