Bots rewrite their own instructions with `bot` when nobody asked

Open
#46 0 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

Nobody has claimed this yet.

Assessment

Difficulty
3/5
Estimated time
1-2 days
Newbie friendliness
58/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Active
Tech stack
javascript

Research direction

Start by locating the bot tool description and the bench that records tool calls in steps; the issue does not name their files or tests. Reproduce the unasked instruction rewrite in the side-question or memory-recall runs, then check that the tool description limits instruction changes to explicit requests and that an unasked call appears in steps.

Written by the indexing model from the issue text.

Description

On the bench at Luna low, Kira called bot to rewrite her own instructions in 3 of 6 side-question and deep-recall runs. Each time it was after an ordinary message like "Facts follow, just read them" or a side question. In one memory-recall try, the bot saved a locker code into its instructions instead of relying on memory.

Right looks like: a bot changes its own instructions only when a person asks it to, and the bot tool description says so. The bench counts it: an unasked bot call shows up in steps.

Dominant language
JavaScript
Stars
0
Forks
0
Avg merge
23m
Merged PRs (30d)
8

Getting set up

This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from oskarrough/agent-ide

All issues in oskarrough/agent-ide

Similar issues

More JavaScript issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.