Toolkit
Persistent Memory · Compositions
On this page6 sections
Compositions
Memory + Project Instructions
Instructions hold stable team contracts; memory holds learned or changing context. Let instructions route to memory without allowing memory to override policy silently.
Memory + Evidence
Store pointers to verified artifacts rather than unsupported conclusions. Recompute or re-open the evidence when the decision matters.
Memory + Subagents
Role memory can preserve specialist calibration. Shared decisions belong in a common reviewed source so workers do not accumulate incompatible context.
Memory + Hooks
Hooks may capture candidate learnings or freshness signals. Keep model-written updates reviewable and prevent recursive or unbounded writes.
Memory + MCP
Use memory to remember stable identifiers and query routes. Ask the authoritative external system for current state; never store credentials.
Memory + Knowledge Governance
For organizational use, add ownership, classification, retention, access control, correction history, and deletion policy. The result is not merely “better prompting”; it is a managed information system.
Source register
Persistent Memory sources
Primary vendor documentation reviewed on the date shown. Links may change after publication.
- How Claude remembers your project (opens in a new tab)
Anthropic · official docs · accessed 2026-08-01
- Custom instructions with AGENTS.md (opens in a new tab)
OpenAI · official docs · accessed 2026-08-01
- Memories (opens in a new tab)
OpenAI · official docs · accessed 2026-08-01
- GitHub Copilot CLI command reference (opens in a new tab)
GitHub · official docs · accessed 2026-08-01