Development

Skill

/evolve-lite-provenance

Analyze saved trajectories and recall audit events offline to record whether recalled guidelines influenced completed sessions.

Boost this skill

agenttoolkit@agenttoolkit

GitHub

From plugin

altk-evolve

11631 skills11 commands

Install

$ npx -y skills add AgentToolkit/altk-evolve --skill evolve-lite-provenance --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.
You can call itInvoke it directly when you want it.
Slash command/evolve-lite-provenance

Context preview

The summary Claude sees to decide when to auto-load this skill.

Analyze saved trajectories and recall audit events offline to record whether recalled guidelines influenced completed sessions.

SKILL.md

evolve-lite-provenance.SKILL.md

name: evolve-lite:provenance
description: Analyze saved trajectories and recall audit events offline to record whether recalled guidelines influenced completed sessions.

Provenance Analyzer

Overview

This skill runs after one or more sessions have completed. It reads `recall` events from `.evolve/audit.log`, locates each session's trajectory, and records post-hoc `influence` events for the recalled guidelines.

The mechanical work — reading recall rows, skipping already-assessed pairs, resolving entity files, and locating trajectories — is done deterministically by `provenance.py candidates`. Your job is the judgment: read each candidate and decide whether the recalled guideline was `followed`, `contradicted`, or `not_applicable`, then persist that verdict.

Use this skill when you want to compute usage provenance without coupling the work to the live learn step.

Workflow

Step 1: Get candidates

Run the candidate builder. It emits one JSON object per line (JSONL), one per unresolved `(session_id, entity)` recall pair:

python3 .bob/skills/evolve-lite-provenance/scripts/provenance.py candidates

Each candidate looks like:

{
  "session_id": "<session-id>",
  "entity_id": "<type>/<name>",
  "entity_excerpt": "<frontmatter + content of the entity file>",
  "trajectory_path": "/path/to/transcript.jsonl",
  "trajectory_excerpt": "<head of the trajectory transcript>",
  "missing": ["trajectory"]
}

Notes:

`entity_id` is the path relative to `.evolve/entities/` without the `.md`

suffix, e.g. `feedback/foo`, `guideline/bar`, or `subscribed/alice/guideline/baz`.

Pairs that already have an `influence` row are skipped for you — the builder

reuses the same dedup rule used when influence rows are written. You will never be handed a duplicate.

The trajectory locator checks `.evolve/trajectories/` first, then falls back

to the native Claude transcript at `~/.claude/projects/<slug>/<session-id>.jsonl`. This means provenance works even when no `.evolve/trajectories/` file was written.

If an entity file or trajectory cannot be found, the candidate is still

emitted with a `missing: [...]` field so the gap is visible. When the trajectory is missing you usually cannot judge the pair — skip it (do not guess), unless the entity content alone makes `not_applicable` certain.

Step 2: Judge each candidate

For each candidate, read `entity_excerpt` (and open `trajectory_path` for the full transcript if the excerpt is not enough). Compare the recalled guideline against the agent's actual actions in the trajectory and pick exactly one verdict:

`followed` — the agent's actual actions are consistent with the guideline.
`contradicted` — the guideline applied, but the agent did the opposite or

repeated the avoidable dead end.

`not_applicable` — the guideline was recalled but did not apply to this

session.

Keep `evidence` to one short sentence citing a concrete action, tool call, or absence in the trajectory. This judgment is yours — there is no heuristic fallback.

Step 3: Record verdicts

Persist each verdict. Either pipe one verdict per call to `provenance.py record`:

echo '{
  "session_id": "<session-id>",
  "entity": "<type>/<name>",
  "verdict": "followed",
  "evidence": "Agent used the saved parser before trying shell fallbacks."
}' | python3 .bob/skills/evolve-lite-provenance/scripts/provenance.py record

…or, to batch many assessments for one session in a single call, pipe to the underlying writer directly:

echo '{
  "session_id": "<session-id>",
  "assessments": [
    {"entity": "feedback/foo", "verdict": "followed", "evidence": "Agent followed it."},
    {"entity": "guideline/bar", "verdict": "not_applicable", "evidence": "Did not apply."}
  ]
}' | python3 .bob/skills/evolve-lite-provenance/scripts/log_influence.py

Both paths write the identical `influence` audit row and skip duplicates. The `entity` value must match the candidate's `entity_id` exactly, including any `subscribed/<source>/` prefix.

It is valid to record nothing when recall events exist but no recalled guideline can be assessed (e.g. every candidate is missing its trajectory).

Read it on GitHub ↗

Ships withaltk-evolve

Coding agents repeat the same mistakes because they start fresh every session. Evolve gives agents memory — they learn from what worked and what didn't, so each session is better than the last.

Get the whole plugin

Stats

Python Bash PostgreSQL REST API Kubernetes

+6 tags

Git GitHub Anthropic OpenAI Hugging Face MCP

116

Stars

Forks

Active

Maintenance

Python

Language