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/robinslange/learning-loop/dream-evalnpx skills add robinslange/learning-loop --skill dream-evalgit clone --depth 1 https://github.com/robinslange/learning-loopWhat 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 | $0.00081 | $0.00624 |
| Opus 5 | $0.00041 | $0.00312 |
| Sonnet 5 | $0.00016 | $0.00125 |
| Haiku 4.5 | $0.00008 | $0.00062 |
Grade A, and why
dream-eval scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 35 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dream Eval: Measure a Consolidation Pass
Overview
wraps /dream without changing it, and answers three questions with a number: did this pass help, is /dream better than no consolidation at all, and does it degrade the corpus over repeated passes. report-only. control mode is the one that matters most: on a low-redundancy corpus the published prior says consolidation only ties raw retrieval, so proving it locally is the point.
Argument Parsing
| Input | Mode | Passes | Mine first |
|---|---|---|---|
/learning-loop:dream-eval |
single | n/a | no |
/learning-loop:dream-eval --mode=control |
control | n/a | no |
/learning-loop:dream-eval --mode=repeated --passes=5 |
repeated | 5 | no |
/learning-loop:dream-eval --mine |
(mode) | yes |
Steps
- acquire the dream lock first, before any mining or mode run, using Bash:
node "${CLAUDE_PLUGIN_ROOT}/scripts/marker.mjs" lock-acquire dream. exit 0 means proceed. exit 1 means another /dream or dream-eval run is active (or a crashed one less than an hour old), so stop and tell the user. exit 2 means a usage or install error, report the stderr message and abort. - if
probes.jsonlis absent and--minewas not passed, stop and say: "no probe corpus found. run with --mine first." (release the lock before stopping, see step 6). - if
--mine, run the probe miner (forward + reverse) and persist to probes.jsonl. - run the chosen mode. the retrieval function is an in-session Task dispatch: given the question and the MEMORY.md index, pick up to 3 files to read. single mode snapshots the live dir first; control and repeated operate on clones only.
- write the json + markdown report under dream-eval/reports/ and show the markdown summary inline.
- release the dream lock when done, after the report is written, always, using Bash:
node "${CLAUDE_PLUGIN_ROOT}/scripts/marker.mjs" lock-release dream.
Safety
- shares the dream lock (acquired in step 1, released in step 6) so the harness and /dream never run at once.
- single mode mutates the live memory dir (snapshot taken first); control and repeated never touch it.
- archive over delete.
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 · 35 lines · 81 tokens per session scan A e26ed852db88
dream-eval is a skill published in the GitHub repository robinslange/learning-loop (11 stars, last pushed 11d ago), licensed Apache-2.0. It adds 81 tokens to every session and 624 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
agent-host-chat-contributions
Build and review cross-cutting agent-host chat behavior through lifecycle contributions. Use when adding turn lifecycle side effects, prompt or context injection, restored-history transformation, protocol-action observation, or when reviewing changes that add code to AgentSideEffects or AgentService.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.