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 skills add wallneradam/project-mem-mcp --skill dreamgit clone --depth 1 https://github.com/wallneradam/project-mem-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/skills/wallneradam/project-mem-mcp/dream)<a href="https://agentmods.dev/skills/wallneradam/project-mem-mcp/dream"><img src="https://agentmods.dev/badge/skills/wallneradam/project-mem-mcp/dream.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 5.1 | $0.00144 | $0.02010 |
| Opus 5 | $0.00072 | $0.01005 |
| Sonnet 5 | $0.00029 | $0.00402 |
| Haiku 4.5 | $0.00014 | $0.00201 |
Grade A, and why
dream 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 8d 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 — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dream — Project Memory Consolidation
When this skill is triggered, invoke the dream-consolidator named subagent (defined in plugin/agents/dream-consolidator.md, pinned to model: claude-sonnet-5) to consolidate the project's MEMORY.md.
Protocol
-
Read all CLAUDE.md files in the project (
**/CLAUDE.md). -
Consolidate by calling the
Agenttool withsubagent_type: "project-mem:dream-consolidator"(the fully-qualifiedplugin:agentname — the baredream-consolidatoris NOT found and the call fails). Pass the inputs in theprompt. Do NOT pass amodel:parameter — see Important.The
promptis a CLOSED TEMPLATE. Send exactly these fields, in this order, and nothing else:Today's date: {TODAY} Project path: {CLAUDE_PROJECT_DIR} ## Preserve verbatim — added or revised in this session: - {pointer} - {pointer} ## Project CLAUDE.md files: ### {path/to/CLAUDE.md} {content} ### {path/to/another/CLAUDE.md} {content}Substitute
{TODAY}with today's date inYYYY-MM-DDformat and{CLAUDE_PROJECT_DIR}with the absolute project root before sending. Do NOT inline MEMORY.md — the subagent loads it itself viaget_project_memory(with chunked reads if large; see its Read protocol). The agent's body (its system prompt) already contains the consolidation rules and Recent Sessions retention policy — do not duplicate them in the user prompt.
The Preserve verbatim field — the ONLY place to protect content
If this session wrote knowledge into MEMORY.md that must survive the consolidation untouched, list it here. Everything you might otherwise be tempted to add as free prose ("be conservative", "keep the measured numbers", "this is important") belongs in this field or nowhere.
The field exists because it fills a real gap, not just to keep you tidy: the consolidator's dormant-topic heuristic keys off the last ## Recent Sessions mention, so a body section written an hour ago — one that Recent Sessions may not name — can read as dormant and get collapsed. A pointer here makes it immune.
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.
- 8d ago First seen · 71 lines · 144 tokens per session scan A 5809aedfbb61
dream is a skill published in the GitHub repository wallneradam/project-mem-mcp (2 stars, last pushed 20d ago), licensed MIT. It adds 144 tokens to every session and 2,010 once invoked, about $0.0007 per session on Opus 5. 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 2026-08-31.
Other skills, from other repositories
memory-bank
Token-efficient persistent memory system for Claude Code that saves 67% tokens on session warm-up (verified with tiktoken). Layered architecture with progressive loading, compact encoding, branch-aware context, smart compression, session diffing, conflict detection, session continuation protocol, and recovery mode.…
claude-bridge-role-memory-keeper
Playbook for a Claude Code agent in the dedicated memory keeper role — single-writer for shared agent memory across a team of 3+ peers. Use when you are designated as the keeper (= other peers send you write candidates, you write them and watch for drift). Memory hygiene + reconciliation against the canon + drift…
memai-maintenance
Orchestrator of the MemAI MAINTENANCE pass (long-term agent memory). Runs [[memai-checkpoints]], [[memai-distill]], [[memai-confidence]] and [[memai-link]] as one curation pass, owns what none of them covers — the DOMAIN TREE (nested paths, cross-listing, casing), the DIAGRAMS and the store's DECAY (reviewafter) — and…
memai-confidence
Verify a MemAI memory's claims against live facts and set its confidence (unverified | confirmed | contradicted). Checks the anchors a scan already extracted — file:line, symbol, table/column, domain, URL — against today's source, promotes or demotes each memory, stages a reword for a detail that drifted, and pushes…
memai-checkpoints
The lifecycle of MemAI CHECKPOINTS: deciding which bearing is still current and which one is superseded, archiving the checkpoint of work that is closed or an intra-session trail that a newer one already covers, keeping the single bearing a live effort still needs, correcting a fact in a kept checkpoint that has gone…
memai-distill
Distill and MOVE knowledge between MemAI types: pull what is durable out of one checkpoint (or several) into a note/reasoning/antipattern, change a memory's type (checkpoint→note when the content is timeless), and merge near-duplicates. Use when the user asks to "distill memory", "turn a checkpoint into a note", "move…