Skip to content

docs: clarify task-agent context boundaries#2152

Closed
NubsCarson wants to merge 4 commits into
developfrom
nubs/agent-context-docs-20260515
Closed

docs: clarify task-agent context boundaries#2152
NubsCarson wants to merge 4 commits into
developfrom
nubs/agent-context-docs-20260515

Conversation

@NubsCarson

@NubsCarson NubsCarson commented May 15, 2026

Copy link
Copy Markdown
Contributor

Summary

Refreshes the task-agent context documentation in AGENTS.md on top of current develop.

  • Clarifies that Codex agents read the workspace AGENTS.md plus orchestrator-generated task context.
  • Separates stable project guidance from runtime character/config policy.
  • Documents that Claude Code project assets under .claude/agents/, .claude/commands/, .claude/plugins/, and .claude/hooks/ are out-of-band for Codex instructions.

Sync Status

  • Merged latest origin/develop through faddd9390 (#2161).
  • Kept the PR docs-only: AGENTS.md.
  • Branch head: fa14282b6.

Validation

No runtime validation required; this is documentation-only.

Local validation:

git diff --check

The branch was refreshed with a normal merge, not rebased, to avoid rewriting prior branch history.

@macroscopeapp macroscopeapp Bot requested a review from Dexploarer May 15, 2026 00:19
@macroscopeapp

macroscopeapp Bot commented May 15, 2026

Copy link
Copy Markdown

Approvability

Verdict: Needs human review

Unable to check for correctness in fa14282. Documentation-only changes with no runtime impact, but AGENTS.md is owned by @elizaos/core-maintainers and the author is not a designated owner of this file.

You can customize Macroscope's approvability policy. Learn more.

@github-actions

Copy link
Copy Markdown
Contributor

Credit balance is too low

@github-actions github-actions Bot added category:docs Auto-managed semantic PR category trust:probationary Building trust, closer scrutiny (auto-managed) labels May 15, 2026
@github-actions

Copy link
Copy Markdown
Contributor

Credit balance is too low

@github-actions github-actions Bot added category:docs Auto-managed semantic PR category trust:probationary Building trust, closer scrutiny (auto-managed) and removed trust:probationary Building trust, closer scrutiny (auto-managed) category:docs Auto-managed semantic PR category labels May 19, 2026
@NubsCarson

Copy link
Copy Markdown
Contributor Author

Base workflow unblocker split out as #2157: #2157. That PR only guards the Auth P0 local-Eliza prebuild when package-mode checkouts do not have eliza/packages/core. Once that lands on develop, subsequent pull_request_target Auth P0 runs should stop failing before the real auth test suite.

@NubsCarson

Copy link
Copy Markdown
Contributor Author

Correction on #2157: I closed the standalone guard PR because it cannot independently pass current package-mode CI from bare develop; the typecheck/build failures require the package-mode UI/runtime fixes already bundled in #2153. The Auth P0 guard remains in #2153, and #2153 is the real combined unblocker for this area.

@NubsCarson

Copy link
Copy Markdown
Contributor Author

Status refresh (2026-05-19): no source changes pushed here. This docs PR is still blocked only by the shared Auth P0 / Agent Review pull_request_target package-mode issue; the previously mentioned standalone #2157 was closed, and the current workflow/base guard follow-up is #2158 stacked after #2153. No new device/runtime claims are added by this refresh.

@codacy-production codacy-production Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull Request Overview

This PR currently contains no file changes, which prevents the verification of the intended documentation updates regarding task-agent context boundaries. All identified acceptance criteria—including clarifying Codex agent context consumption, separating project guidance from runtime policy, and documenting the exclusion of the .claude/ directory—remain unaddressed. As there are no code changes, no logic or security issues were found, but the PR cannot be merged in its current state.

About this PR

  • The pull request contains no file changes. The summary indicates that updates to AGENTS.md are required to clarify Codex context boundaries and .claude/ directory exclusions. Please commit the missing documentation changes.

Test suggestions

  • Documentation of Codex agent context consumption (AGENTS.md + orchestrator)
  • Distinction between static project guidance and runtime policy
  • Documentation of .claude/ directory exclusion
Prompt proposal for missing tests
Consider implementing these tests if applicable:
1. Documentation of Codex agent context consumption (AGENTS.md + orchestrator)
2. Distinction between static project guidance and runtime policy
3. Documentation of .claude/ directory exclusion

TIP Improve review quality by adding custom instructions
TIP How was this review? Give us feedback

@codacy-production

codacy-production Bot commented May 22, 2026

Copy link
Copy Markdown

Not up to standards ⛔

🔴 Issues 1 minor

Alerts:
⚠ 1 issue (≤ 0 issues of at least minor severity)

Results:
1 new issue

Category Results
Comprehensibility 1 minor

View in Codacy

AI Reviewer: first review requested successfully. AI can make mistakes. Always validate suggestions.

Run reviewer

TIP This summary will be updated as you push new changes.

@github-actions

Copy link
Copy Markdown
Contributor

Credit balance is too low

@github-actions github-actions Bot added category:docs Auto-managed semantic PR category trust:probationary Building trust, closer scrutiny (auto-managed) and removed trust:probationary Building trust, closer scrutiny (auto-managed) category:docs Auto-managed semantic PR category labels May 25, 2026
Comment thread AGENTS.md
Milady is a local-first AI assistant built on [elizaOS](https://github.com/elizaOS). It wraps the elizaOS runtime with a CLI, desktop app (Electrobun), web dashboard, and platform connectors (Telegram, Discord, WeChat, etc.). Inherits `~/.codex/AGENTS.md`.
Milady is a local-first AI assistant built on [elizaOS](https://github.com/elizaOS). It wraps the elizaOS runtime with a CLI, desktop app (Electrobun), web dashboard, and platform connectors (Telegram, Discord, WeChat, etc.).

Codex agents read this `AGENTS.md` from the workspace. When spawned by the agent orchestrator, they also receive generated task context with the current room, task, repo routes, and loopback bridge endpoints. Keep stable project rules here; keep deployment/persona policy in the runtime character/config.

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Codacy has a fix for the issue: Overly complex sentence (43 words). Break into shorter instructions.

Suggested change
Codex agents read this `AGENTS.md` from the workspace. When spawned by the agent orchestrator, they also receive generated task context with the current room, task, repo routes, and loopback bridge endpoints. Keep stable project rules here; keep deployment/persona policy in the runtime character/config.
Split into multiple lines. Config files should read like command lists, not paragraphs.

@lalalune

Copy link
Copy Markdown
Contributor

Closing as part of the Milady -> elizaOS/eliza PR migration sweep. This docs-only PR modifies Milady-specific AGENTS.md context and fails existing checks; the current upstream AGENTS.md already carries the active contributor charter for elizaOS, so there is no clean upstream migration patch.

@lalalune lalalune closed this May 29, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

category:docs Auto-managed semantic PR category trust:probationary Building trust, closer scrutiny (auto-managed)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants