Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/spideynolove/memory/reflectgit clone --depth 1 https://github.com/spideynolove/memoryWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00013 | $0.01335 |
| Opus 5 | $0.00006 | $0.00668 |
| Sonnet 5 | $0.00003 | $0.00267 |
| Haiku 4.5 | $0.00001 | $0.00134 |
Grade B, and why
reflect scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
cat /home/hung/MCPs/memory/.claude/reflections/processed.log 2>/dev/null || echo "(empty)" How it starts
The opening of the file, as written. The whole thing — 176 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reflect on Diary Entries and Synthesize Insights
Analyze diary entries from .claude/diary/ to identify recurring patterns, synthesize insights, and propose updates to the project's CLAUDE.md.
Storage
- Diary entries:
<project-root>/.claude/diary/ - Reflection outputs:
<project-root>/.claude/reflections/ - Processed log:
<project-root>/.claude/reflections/processed.log - Target config:
<project-root>/CLAUDE.md(project-specific rules)
Project root: /home/hung/MCPs/memory
Parameters
- Entry count: "last N entries" (default: last 10)
- Date range: "from YYYY-MM-DD to YYYY-MM-DD"
- Pattern filter: "related to [keyword]"
Steps
1. Check processed entries log
cat /home/hung/MCPs/memory/.claude/reflections/processed.log 2>/dev/null || echo "(empty)"
Format: [diary-filename] | [reflection-date] | [reflection-filename]
2. Locate unprocessed diary entries
ls -t /home/hung/MCPs/memory/.claude/diary/*.md 2>/dev/null | head -20
Exclude entries already listed in processed.log unless user says "include all".
3. Read and parse diary entries
Read each entry. Pay special attention to:
- User Preferences Observed
- Code Patterns and Decisions
- Solutions Applied
- Challenges Encountered
- Code Review & PR Feedback
4. Optional: query Memory Bank MCP for additional signal
If the Memory Bank MCP server is running, also pull structured patterns:
- Call
get_analysisresource (memory://analysis) to get detected patterns from SQLite - Call
get_memory_treeresource (memory://tree) for structured session memories
This hybrid approach combines narrative signal (diary markdown) with structured signal (SQLite patterns). Skip gracefully if MCP is unavailable.
5. Read current CLAUDE.md
cat /home/hung/MCPs/memory/CLAUDE.md 2>/dev/null || echo "(no CLAUDE.md yet)"
This is critical — check for existing rules before proposing additions or strengthening.
6. Flag correction-priority entries
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 176 lines · 13 tokens per session scan B 925340b16139
reflect is a command published in the GitHub repository spideynolove/memory (3 stars, last pushed 5mo ago), licensed MIT. It adds 13 tokens to every session and 1,335 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
handoff
Update memory-bank/handoff.md to pause or switch models.
save-context
Write a phase-completion summary and update the Memory Bank.
start-task
Plan a task with minimal context before coding.
add-project
在 Obsidian vault 里新增一个项目。自动检测 vault（不存在则 inline 初始化）、分析代码结构生成架构草稿、写入 CLAUDE.md。通常是接触新项目时唯一需要运行的命令。.
update-memory
全量对账 Code ↔ Obsidian ↔ auto-memory 三方，找漂移并让用户逐项决定如何修。地图 vs 地形原则的工程兜底。.
commit
智能 git 提交：自动生成 commit message、提交代码、然后调用 /sync-docs 同步到 Obsidian。.