Agent Memory Bridge

Agent Memory Bridge

When corrections, context, or patterns learned in one agent (Hermes, Claude Code, Codex) need to propagate to all others.

When to use

• User says something like "remember this" or gives a correction in Hermes that should apply to Claude Code
• You discover a workspace convention or pattern that other agents on other machines need
• Setting up a new machine to match existing agent context
• After an adversarial review reveals one agent knew something another didn't

Locations

Skill Accessibility Across All Agents

Each agent accesses skills differently. A skill in .claude/skills/ must also be visible through each agent's mechanism. This was verified in #1949 (Codex symlink fix).

CRITICAL VERIFICATION: Run this after any skill changes to confirm all agents can see skills:

echo "=== Agent Skill Accessibility ==="
echo "CC:    $(find -L .claude/skills -name 'SKILL.md' -not -path '*/_archive/*' | wc -l)"
echo "Codex: $(find -L .codex/skills -name 'SKILL.md' -not -path '*/_archive/*' | wc -l)"
echo "Gemini: $(find -L .gemini/skills -name 'SKILL.md' -not -path '*/_archive/*' | wc -l)"

Key Pitfall: .codex/skills/ should be a symlink to ../.claude/skills/ but was accidentally a real directory with only 57 GSD skills (out of 696+). Always verify it's a symlink: ls -la .codex/skills should show -> ../.claude/skills, not a directory.

Sub-repo Behavior: When working in sub-repos (CAD-DEVELOPMENTS/, digitalmodel/, etc.), each agent sees that repo's local .claude/skills/ (31-261 skills) PLUS the workspace-hub skills (696) depending on working directory context. See #1951 for sub-repo skill visibility gap.

Bridging Hermes → Claude Code

1. Hermes memory is always available in the system context — extract the consolidated facts
2. Write to ~/.claude/CLAUDE.md (global) or ~/.claude/projects/<hash>/memory/ (project-scoped)
3. The GLOBAL CLAUDE.md is the single source of truth — write conventions, paths, user preferences, corrections
4. Project-scoped memory is for organic learnings accumulated over sessions

Bridging Claude Code → Hermes

1. Read ~/.claude/projects/*/memory/*.md — these are Claude's learned corrections and context
2. Feed into delegate_task context for subagents
3. Use memory tool to store critical corrections in Hermes memory (curate — Hermes memory is limited to 2200 chars)

Bridging to Licensed Machines (Windows)

1. Export unified memory files to a tarball or git-tracked directory
2. Copy to Windows machine
3. Bootstrap with .claude/CLAUDE.md at project level
4. Use python not uv run on Windows

Key Principles

• Compact source of truth: One global CLAUDE.md that any agent can load
• Curate aggressively: Claude's auto-memory grows organically and has duplicate/cross-cutting entries. Deduplicate before bridging.
• Git-track everything: Put the bridge scripts and export files in the repo so they survive machine loss
• No agent loses context: If Claude's OAuth session dies and starts fresh, it should still have the same baseline knowledge via CLAUDE.md

Pitfalls

• Claude's auto-memory directory is at ~/.claude/projects/-<path>-hash/memory/ — the directory name is URL-encoded path with dashes, not a UUID
• Claude's auto-memory has a 25KB/200-line limit per project
• Multiple project directories exist in parallel (workspace-hub, digitalmodel, worldenergydata) — bridge ALL of them
• Windows path separators differ — test the bootstrap script on the target machine