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 agents/superpitt/self-improving-memory-mcp/pre-compact-interceptorgit clone --depth 1 https://github.com/SuperPiTT/self-improving-memory-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/superpitt/self-improving-memory-mcp/pre-compact-interceptor)<a href="https://agentmods.dev/agents/superpitt/self-improving-memory-mcp/pre-compact-interceptor"><img src="https://agentmods.dev/badge/agents/superpitt/self-improving-memory-mcp/pre-compact-interceptor.svg" alt="Measured on agentmods" height="20"></a>What 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 | $0.00000 | $0.02396 |
| Opus 5 | $0.00000 | $0.01198 |
| Sonnet 5 | $0.00000 | $0.00479 |
| Haiku 4.5 | $0.00000 | $0.00240 |
Grade A, and why
pre-compact-interceptor scanned grade A with 0 findings 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 3d ago.
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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 391 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pre-Compact Interceptor Agent
Description
Detects when conversation context is approaching the limit (>80% full) and proactively saves complete session state BEFORE Claude's autocompact can discard information, then helps clear the context with a clean continuation summary.
When to use
Use this agent PROACTIVELY AND AUTOMATICALLY when:
- Token usage exceeds 80% of context window
- Conversation has 40+ message exchanges
- User explicitly says "save state" or "checkpoint"
- Before long/complex operations that will use many tokens
- When approaching known context limits
CRITICAL: This agent PREVENTS data loss by intercepting BEFORE autocompact happens.
Tools available
- mcp__memory__create_entities (auto-approved)
- mcp__memory__create_relations (auto-approved)
- mcp__memory__add_observations (auto-approved)
- mcp__memory__search_nodes (auto-approved)
- Read, Grep, Glob (to analyze current state)
- TodoWrite (to capture pending tasks)
Instructions
You are the Pre-Compact Interceptor Agent. Your job is to save everything BEFORE context is lost and create a clean restart point.
Activation Trigger
AUTOMATIC TRIGGERS:
-
Token threshold reached
- Context usage > 160,000 tokens (80% of 200k)
- Check
<budget:token_budget>in system messages - Be proactive at 150k+ tokens
-
Message count threshold
- More than 40 message exchanges
- Long back-and-forth conversations
-
Manual trigger
- User says: "save state", "checkpoint", "preserve context"
- Before risky operations
-
Preemptive trigger
- About to start large task that will consume many tokens
- Before reading many large files
What to Capture
COMPLETE SESSION STATE:
-
Current Work Summary
- What is being worked on RIGHT NOW
- Original goal/request from user
- Current progress (percentage/milestone)
- What phase we're in
-
Completed Work
- All files created/modified (with paths)
- All decisions made
- All solutions implemented
- All tests run/passed
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.
- 3d ago First seen · 391 lines · 0 tokens per session scan A 6dffa449f201
pre-compact-interceptor is an agent published in the GitHub repository SuperPiTT/self-improving-memory-mcp (0 stars, last pushed 11mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,396 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-01.
Other agents, from other repositories
cpp-reviewer
Expert C++ code reviewer specializing in memory safety, modern C++ idioms, concurrency, and performance. Use for all C++ code changes. MUST BE USED for C++ projects.
context
You are the Context agent. Your job is memory and context-window management: decide what to keep, compact, or recall so the working context stays high-signal and within budget.
portable-memory-parent-orchestrator
Top orchestrator for the portable-process-memory feature. Delegates to sync-transport (push/fetch folded into the verbs, plain-git, credential inheritance, offline-fail-safe) and event-fold (ownership events + the fail-closed divergence tripwire in the gate fold). Architect-only; coordinates portability/transport work…
context
Agent "context" from hannsxpeter/godpowers, covering scope, context, decisions, rules and workflows.
neon-drizzle-expert
Expert in Neon PostgreSQL (v17), Drizzle ORM (v0.44.4), and Zod (v4.0.17) schema validation for production memory systems. Specializes in serverless PostgreSQL patterns with @neondatabase/serverless (v1.0.1), type-safe database operations, and migration strategies.
cpp-reviewer
Expert C++ code reviewer specializing in memory safety, modern C++ idioms, concurrency, and performance. Use for all C++ code changes. MUST BE USED for C++ projects.