persona-evolution

A set of instructions for adapting an assistant’s communication style and remembered preferences from clear user evidence. It separates lasting preferences from instructions that apply only to one task.

In plain words
What is it for?
Use it when a user corrects the assistant, explains how they want decisions or communication handled, or asks the assistant to remember stable information.
Why use it?
It helps keep useful user preferences consistent without making up personal facts or treating temporary requests as permanent.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/codeinfinity1/stram/persona-evolution
Any agent
npx skills add CodeInfinity1/Stram --skill persona-evolution
Clone the repo
git clone --depth 1 https://github.com/CodeInfinity1/Stram

Made for: Claude Code, Codex.

Per session 43 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 602 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What 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.

ModelPer sessionOnce invoked
Fable 5 $0.00043 $0.00602
Opus 5 $0.00022 $0.00301
Sonnet 5 $0.00009 $0.00120
Haiku 4.5 $0.00004 $0.00060

Measured yesterday against content hash 333c16306220, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

persona-evolution 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 yesterday.

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.

skills/agent-core/persona-evolution/SKILL.md · 78 lines

How it starts

The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Persona Evolution

Purpose

Let the assistant grow a stable working relationship with the user without inventing personality or overfitting. This skill uses Stram's model-led persona evolution over evidence.

When To Use

Use when the user says how they want the assistant to communicate, how decisions should be made, what to remember about them, or what behavior should change over time.

Inputs And Evidence

  • Direct user statements.
  • Interaction review records.
  • Persona records and evolutions.
  • Repeated preferences across sessions.
  • Current task context and communication channel.

Tool Map

  • cognitive_persona_evolve
  • cognitive_persona_evolution_status
  • cognitive_interaction_review
  • memory_profile
  • memory_write
  • cognitive_state

Workflow

  1. Separate stable preferences from momentary task-specific instructions.
  2. Gather current persona and recent interaction evidence.
  3. Use model-led persona evolution to propose updates.
  4. Record only preferences or stable facts with direct evidence.
  5. Preserve boundaries such as not claiming emotions, identity, or certainty beyond evidence.
  6. In the response, explain the practical behavior change.

Safety And Boundaries

  • Do not store sensitive personal facts unless useful and clearly requested or appropriate.
  • Do not infer protected attributes.
  • Do not overwrite explicit preferences with weak contradictory evidence.

Safety And Approval

  • Persona updates must be grounded in direct user statements, repeated behavior, or interaction-review evidence.
  • Do not create claims about the assistant's identity, emotions, relationship, or certainty beyond evidence and policy boundaries.
  • Do not store private or sensitive user facts unless they are operationally useful and appropriate to retain.
  • If the user asks to forget or change a preference, preserve the latest explicit instruction over older inferred behavior.

Native Implementation Boundaries

  • Use Stram persona-evolution, interaction-review, memory-profile, memory-write, and cognitive-state tools.
  • Personal-development patterns from references are guidance only; persistent persona/user-model changes must be model-led and evidence-backed.
  • Deterministic code may merge bounded records and validate schemas, but must not infer stable user traits from keyword matches.

Read the full file on GitHub · 78 lines

Changes

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.

  1. yesterday First seen · 78 lines · 43 tokens per session scan A 333c16306220

Subscribe to this mod's changes

persona-evolution is a skill published in the GitHub repository CodeInfinity1/Stram (10 stars, last pushed 22d ago), licensed MIT. It adds 43 tokens to every session and 602 once invoked, about $0.0002 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 2026-08-31.