Closed
Conversation
- Added agent_id parameter to AxClient (sends X-Agent-Id header) - Added resolve_agent_id() to config (AX_AGENT_ID env or config.toml) - Updated CLAUDE.md with identity model documentation - Agent-bound PATs with header now act as the agent principal Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
keys create: - Added --agent flag (resolves name to UUID) - Added docstring with examples for bound vs unrestricted PATs - Shows "Bound to: <uuid>" on creation auth init: - Added --agent-id flag for agent-bound PAT config Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…gent Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Design spec covers: - ax listen --exec (agent wake-up primitive) - ax send with SSE reply-waiting (replaces polling) - ax monitor alias - Shared SSEStream connection layer with reconnect and dedup Implementation plan: 11 TDD tasks covering SSE parser, dedup tracker, SSEStream class, mention detection, exec runner, listen command, send SSE integration, and end-to-end verification. Approved by @orion. Reviewed and revised after code review. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add two SSE-based scripts that demonstrate how agents connect to aX: - ax_listener.py: mention monitor — listens for @mentions via SSE, responds via REST API. Includes dedup, loop prevention, backoff. - echo_agent.py: minimal echo bot — receives any message via SSE, echoes it back as a threaded reply. Proof of the SSE→API loop. Both use config from .ax/config.toml, support --dry-run and --verbose. Key discovery: API silently blocks self-mentions (returns 200 empty body). Documented in both scripts. CLAUDE.md updated to wire_tap identity. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…valuation Comprehensive spec for expanding ax-cli from basic API wrapper to the definitive agent toolkit. Concierge-first design: ax send routes through aX by default, ax check gets tailored briefings with reason. Includes live endpoint evaluation (2026-03-18) testing every backend endpoint with the swarm PAT. Documents what works (55+ endpoints), backend bugs (4), auth limitations, and non-existent endpoints to skip. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
After bind, agent_id is the authoritative identity header. agent_name is now limited to bootstrap (no id yet) and explicit interactive --agent targeting. Root cause: resolve_agent_id() returned None when agent_name was present, causing bound PATs to send X-Agent-Name instead of X-Agent-Id. Backend rejected this with 403 on agent-scoped PATs when name resolution failed or was ambiguous. Changes: - config.py: simplify resolve_agent_id() to always return id if available; get_client() only resolves name when no id exists - client.py: flip _agent_headers() precedence — id wins over name - messages.py: _configure_send_identity() preserves default identity when no explicit flags are passed (fixes 403 on scoped PATs) - tests: 16 tests covering id-first precedence, bootstrap fallback, env var priority, per-request overrides, and send identity Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
91d6b04 skipped resolve_agent_name() entirely when config had agent_id, making env-level name targeting impossible. Now AX_AGENT_NAME env beats config-level agent_id (explicit targeting override), while both env vars set still prefers ID as canonical. Precedence: 1. AX_AGENT_NAME env (no AX_AGENT_ID) → name (targeting) 2. AX_AGENT_ID env → ID (canonical) 3. config agent_id → ID (steady state) 4. config agent_name → name (bootstrap) Added regression tests: - env name overrides config id - both env vars set → id wins Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Member
Author
|
Closing — SVGs cherry-picked into PR #9 (clean branch). The rest of the changes in this PR are unrelated. Thanks for the diagrams. |
7 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
$## Summary\n- add three dark-theme SVG diagrams for the axctl README\n- cover platform overview, supervision loop, and profile fingerprint flow\n- keep layouts sized for GitHub README rendering\n\n## Validation\n- opened the SVG files locally and checked XML parses cleanly