Evolve Skill From Traces

Evolve SKILL.md files from agent execution traces using a three-stage pipeline: trajectory collection from observed runs, parallel multi-agent patch proposal for error and success analysis, and conflict-free consolidation of overlapping edits via prevalence-weighting. Based on the Trace2Skill methodology.

45 stars 4 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add pjt222/agent-almanac --skill evolve-skill-from-traces
Manual install
git clone --depth 1 https://github.com/pjt222/agent-almanac.git /tmp/agent-almanac
mkdir -p ~/.claude/skills
cp -r /tmp/agent-almanac/i18n/caveman-lite/skills/evolve-skill-from-traces ~/.claude/skills/evolve-skill-from-traces

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Evolve a Skill from Execution Traces

Transform raw agent execution traces into a validated SKILL.md through a three-stage pipeline: trajectory collection, parallel multi-agent patch proposal, and conflict-free consolidation. This skill bridges the gap between observed agent behavior and documented procedures, turning successful runs into reproducible skills.

## When to Use

- Execution traces reveal recurring patterns not captured in existing skills
- Observed agent behavior outperforms the documented procedure
- Building skills from scratch by recording expert demonstrations
- Multiple agents propose conflicting improvements to the same skill

## Inputs

- **Required**: `traces` -- set of agent execution logs or session transcripts (minimum 10 successful runs recommended)
- **Required**: `target_skill` -- path to an existing SKILL.md to evolve, or `"new"` for skill extraction from scratch
- **Optional**: `analyst_count` -- number of parallel analyst agents to spawn (default: 4)
- **Optional**: `held_out_ratio` -- fraction of traces reserved for validation, not used in drafting (default: 0.2)

## Procedure

### Step 1: Collect Execution Traces

Gather agent session logs, tool-call sequences, or conversation transcripts that demonstrate the target behavior. Filter for runs tagged as successful. Normalize into a standard trace format: a sequence of (state, action, outcome) triples with timestamps.
Read the full skill on GitHub

Details

Skill name
evolve-skill-from-traces
Path
i18n/caveman-lite/skills/evolve-skill-from-traces/SKILL.md
License
MIT
Repo updated
18 November 2026

FAQ

What does the Evolve Skill From Traces skill do?

Evolve SKILL.md files from agent execution traces using a three-stage pipeline: trajectory collection from observed runs, parallel multi-agent patch proposal for error and success analysis, and conflict-free consolidation of overlapping edits via prevalence-weighting. Based on the Trace2Skill methodology.

How do I install the Evolve Skill From Traces skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add pjt222/agent-almanac --skill evolve-skill-from-traces. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Evolve Skill From Traces popular?

It lives in pjt222/agent-almanac, which has 45 GitHub stars and 4 forks, and is included in 1 curated skill list.

Related skills

Adapt Architecture

45

pjt222/agent-almanac

Execute structural metamorphosis using strangler fig migration, chrysalis phases, and interface preservation. Covers transformation planning, parallel running, progressive cutover, rollback design, and post-metamorphosis stabilization for system architecture evolution. Use when assess-form has classified the system as READY for transformation, when migrating from monolith to microservices, when replacing a core subsystem while dependents continue operating, or when any architectural change must be gradual rather than big-bang.

Software DevelopmentIn 1 list

Adaptic

45

pjt222/agent-almanac

Master skill composing the 5-step synoptic cycle for panoramic synthesis across multiple domains. Orchestrates meditate, expand-awareness, observe, awareness, integrate-gestalt, and express-insight into a coherent process that produces unified understanding rather than sequential compromise. Use when a problem genuinely spans 3+ domains and the interactions between domains matter more than depth in any one, when sequential analysis feels like compromise rather than integration, or before major architectural decisions affecting multiple stakeholders.

OtherIn 1 list

Add Puzzle Type

45

pjt222/agent-almanac

Scaffold a new puzzle type across all 10+ pipeline integration points in jigsawR. Creates the core puzzle module, wires it into the unified pipeline (generation, positioning, rendering, adjacency), adds ggpuzzle geom/stat layers, updates DESCRIPTION and config.yml, extends the Shiny app, and creates a comprehensive test suite. Use when adding a completely new puzzle type to the package or following the 10-point integration checklist to ensure nothing is missed end-to-end.

Testing & QAIn 1 list

Fit Drift Diffusion Model

45

pjt222/agent-almanac

Fit cognitive drift-diffusion models (Ratcliff DDM) to reaction time and accuracy data with parameter estimation (drift rate, boundary separation, non-decision time), model comparison, and parameter recovery validation. Use when modeling binary decision-making with reaction time data, estimating cognitive parameters from experimental data, comparing sequential sampling model variants, or decomposing speed-accuracy tradeoff effects into latent cognitive components.

AI, LLMs & AgentsIn 1 list

Evolve Agent

45

pjt222/agent-almanac

Evolve an existing agent definition by refining its persona in-place or creating an advanced variant. Covers assessing the current agent against best practices, gathering evolution requirements, choosing scope (refinement vs. variant), applying changes to skills, tools, capabilities, and limitations, updating version metadata, and synchronizing the registry and cross-references. Use when an agent's skills list is outdated, user feedback reveals capability gaps, tool requirements have changed, an advanced variant is needed alongside the original, or the agent's scope needs sharpening after real

AI, LLMs & AgentsIn 1 list

Fit Hidden Markov Model

45

pjt222/agent-almanac

Fit hidden Markov models using the Baum-Welch (EM) algorithm with model selection, Viterbi decoding for state sequences, and forward-backward probabilities. Use when observations are generated by unobservable latent states, you need to segment a time series into latent regimes (market regimes, speech phonemes, biological sequences), compute sequence probabilities, decode the most likely hidden state path, or compare models with different numbers of hidden states.

AI, LLMs & AgentsIn 1 list

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.