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 skills/cwinvestments/memstack/compressnpx skills add cwinvestments/memstack --skill compressgit clone --depth 1 https://github.com/cwinvestments/memstackWhat 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 | $0.00029 | $0.01157 |
| Opus 5 | $0.00015 | $0.00579 |
| Sonnet 5 | $0.00006 | $0.00231 |
| Haiku 4.5 | $0.00003 | $0.00116 |
Grade A, and why
compress scanned grade A 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 2d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl http://127.0.0.1:8787/health How it starts
The opening of the file, as written. The whole thing — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Compress - TokenStack Proxy Manager
Monitor and troubleshoot the built-in TokenStack compression proxy for CC sessions.
Activation
When this skill activates, output:
Compress - Checking TokenStack status...
Then execute the protocol below.
- Keywords: tokenstack, compression stats, token savings, proxy status, check proxy
- Contextual: When user asks about token usage, context window limits, or session cost
- Level: 2 (explicit trigger only)
Context Guard
| Context | Status |
|---|---|
| User says "tokenstack", "compression stats", "check proxy" | ACTIVE - run status check |
| User asks about token savings or context window | ACTIVE - point to dashboard report |
| Proxy errors or API connection failures appear | ACTIVE - run health diagnostics |
| General discussion about CC features | DORMANT - do not activate |
| User is actively coding (no proxy issues) | DORMANT - do not activate |
What It Does
TokenStack is a transparent proxy between Claude Code and the Anthropic API. It compresses tool output before it reaches the API, extending effective context and lowering token cost.
This skill checks that the proxy is running and routing, and troubleshoots connection issues. For the full feature overview, enable steps, and transform tables, use the Token Optimization skill.
Prerequisites
TokenStack ships inside the memstack-skill-loader package. There is no separate install.
- Start with the dashboard (recommended):
python -m memstack_skill_loader dashboard --with-proxy- Starts the proxy on
127.0.0.1:8787and setsANTHROPIC_BASE_URLautomatically.
- Starts the proxy on
- Run only the proxy:
python -m memstack_skill_loader proxy
Workflow
1. Health Check
curl http://127.0.0.1:8787/health
A healthy proxy responds on /health. If there is no response, the proxy is not running.
Note: the proxy has no /stats endpoint. Live savings are reported in the dashboard, not over curl. See "Reading Savings" below.
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.
- 2d ago First seen · 129 lines · 29 tokens per session scan A 54bf729825e8
compress is a skill published in the GitHub repository cwinvestments/memstack (417 stars, last pushed 5d ago), licensed MIT. It adds 29 tokens to every session and 1,157 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
plan-ceo-review
CEO/founder-mode plan review. Rethink the problem, find the 10-star product, challenge premises, expand scope when it creates a better product. Four modes: SCOPE EXPANSION (dream big), SELECTIVE EXPANSION (hold scope + cherry-pick expansions), HOLD SCOPE (maximum rigor), SCOPE REDUCTION (strip to essentials). Use when…
ship
Ship workflow: detect + merge base branch, run tests, review diff, bump VERSION, update CHANGELOG, commit, push, and create a PR. Use for an explicit /ship invocation or when the user requests the full ship, release, or deploy workflow. For an ordinary commit, push, or pull-request publishing request, use the built-in…
design-review
Designer's eye QA: finds visual inconsistency, spacing issues, hierarchy problems, AI slop patterns, and slow interactions — then fixes them. Iteratively fixes issues in source code, committing each fix atomically and re-verifying with before/after screenshots. For plan-mode design review (before implementation), use…
plan-eng-review
Eng manager-mode plan review. Lock in the execution plan — architecture, data flow, diagrams, edge cases, test coverage, performance. Walks through issues interactively with opinionated recommendations. Use when asked to "review the architecture", "engineering review", or "lock in the plan". Proactively suggest when…
pre-landing-review
Pre-landing PR review. Analyzes diff against the base branch for SQL safety, LLM trust boundary violations, conditional side effects, and other structural issues. Use when explicitly asked for the specialized pre-landing workflow. Product /review requests are handled by BitFun's unified Review mechanism instead.…
autoplan
Auto-review pipeline — reads the full CEO, design, eng, and DX review skills from disk and runs them sequentially with auto-decisions using 6 decision principles. Surfaces taste decisions (close approaches, borderline scope, codex disagreements) at a final approval gate. One command, fully reviewed plan out. Use when…