Catalog
affaan-m/recursive-decision-ledger

affaan-m

recursive-decision-ledger

Use when the user asks for repeated rollouts, marked decision processes, high-dimensional search, stochastic optimization, local-optima exploration, ensemble comparison, or recursive reasoning with a visible evidence trail.

New~603Updated Jul 14, 2026

Recursive Decision Ledger

Use this skill when the user is trying to force deeper computation through repeated rollouts or "Prime Gauss" style recursive prompting. Preserve the useful part: repeated trials, prior memory, fresh information, and explicit marks. Remove the unsafe part: pretending the loop proves certainty.

Ledger Contract

Every rollout should record:

  • rollout id and timestamp;
  • prior accepted winner and prior watchlist;
  • fresh information ingested;
  • search space size;
  • model families or heuristics used;
  • trial count and effective trial count;
  • top candidates;
  • decision marks;
  • coherence marks against the prior ledger;
  • promotion gate result.

Prefer JSONL for append-only ledgers and Markdown for human summaries.

Rollout Loop

  1. Load the prior ledger.
  2. Capture new information at time-step zero.
  3. Run the bounded search.
  4. Mark each candidate: accept, watch, reject, decay watch, or needs replay.
  5. Compare winners against prior winners and latest marked rollout.
  6. Downgrade candidates when drift, tail risk, stale data, or failed replay invalidates the previous mark.
  7. Append artifacts before summarizing.

Coherence Mark

Include a compact coherence mark:

Ensemble matches prior winner: true
Recursive matches prior winner: false
Latest rollout match: true
Live promotion allowed: false
Reason: replay and freshness gates not satisfied

Promotion Rules

For trading, capital allocation, production deploys, migrations, or destructive ops, recursive confidence is not approval.

Default to paper, dry-run, read-only, preview, or staged mode unless the user explicitly approves the live action and the repo/service gate supports it.

Promote only when:

  • the candidate beats the prior accepted winner on the chosen metric;
  • correctness and replay checks pass;
  • risk limits are explicit;
  • the evidence is durable;
  • the user has approved the live step when needed.

Summary Shape

Lead with the decision, not the drama:

Rollout 15 complete. The prior winner still holds, but edge deteriorated 17%.
Status: watch, not live. Next gate: 20 replay fills with fresh orderbook age
below threshold.
Files1
1 files · 1.0 KB

Select a file to preview

Overall Score

76/100

Grade

B

Good

Safety

78

Quality

73

Clarity

77

Completeness

72

Summary

A framework for guiding AI agents through repeated decision cycles with explicit audit trails. The skill structures recursive reasoning as a series of bounded rollouts, each recording prior context, new information, search results, and decision marks. It emphasizes caution in high-stakes actions (trading, deploys, migrations) by defaulting to dry-run/preview modes and requiring explicit user approval before live promotion.

Detected Capabilities

file readfile writebash executiongrep pattern matchingglob file discoverystructured data logging (JSONL)markdown documentation generation

Trigger Keywords

Phrases that MCP clients use to match this skill to user intent.

repeated rollouts optimizationrecursive decision ledgerensemble comparison votingstochastic search with historypromotion gate validationaudit trail decision-making

Risk Signals

WARNING

Skill references trading, capital allocation, production deploys, and migrations as potential use cases

Promotion Rules section
INFO

Skill allows Bash execution without explicit scope guardrails; relies on user discretion and promotion rules

tools field and Promotion Rules section
INFO

Skill emphasizes dry-run and staged modes as default, with explicit user approval required for live actions

Promotion Rules section

Use Cases

  • Run repeated optimization trials with auditable decision history
  • Compare ensemble or recursive models while tracking prior winners
  • Explore high-dimensional search spaces with coherence tracking
  • Make stochastic optimization decisions with explicit evidence trails
  • Gate production changes (trades, deploys, migrations) with multi-rollout validation
  • Replay and validate candidates against fresh data before promoting to live

Quality Notes

  • Skill provides a clear contract (Ledger Contract) that defines what each rollout must record, giving the agent specific structural expectations
  • Coherence Mark section includes a concrete example showing the decision boundary (live promotion allowed: false with explanation)
  • Promotion Rules explicitly default to safe modes (paper, dry-run, read-only, preview, staged) and require user opt-in for live actions—this is strong risk mitigation
  • Summary Shape section demonstrates expected output format with a concrete example
  • Skill is domain-agnostic and applicable to optimization, search, decision-making, and validation workflows
  • The skill appropriately warns against treating recursive confidence as certainty—clarifies the epistemological limits
  • No supporting examples of actual ledger outputs or rollout implementations; feels high-level and abstract
  • Would benefit from concrete code examples showing how to record and compare rollout data in practice
Model: claude-haiku-4-5-20251001Analyzed: Jul 14, 2026

Reviews

Add this skill to your library to leave a review.

No reviews yet

Be the first to share your experience.

Version History

  1. v1.1

    Content updated

    ✦ AINo behavioral changes detected.

    2026-07-14

    Latest
  2. v1.0

    2026-05-25

    View This VersionInitial version

Use affaan-m/recursive-decision-ledger in your dev environment

Command Palette

Search for a command to run...