· 1 observation
Technical updates · 1
Package or technical metadata updated
View technical comparison →Suneel Kaushik S v0.14.0
Coordinate agents on one repo
From the developer
Not specified in the available declarations: limitations, pricing & account requirements.
Codex and I are both free here. Split the open v2.1 issues, claim my half, and tell Codex which ones are its to start.
Developer declarations; not independently tested. Listing last checked Nov 14, 2026. Package observed Nov 14, 2026.
From the marketplace listing
Agent Parley lets several coding agents work on one repository at the same time without overwriting each other. Each participant runs in its own Git worktree. Before editing, a participant claims a numbered issue, and the claim is visible to every other participant. File reservations are announced over MCP and are advisory: peers see a conflict and stop, but nothing on disk is locked. Work moves between participants through an explicit offer and accept, carrying a summary of the commit, the checks that ran, and what remains. Outcomes are reported as agent claims with evidence, never as independent verification. The skill reads coordination state through the agent-parley command line tool, which must be installed separately.
Language: English · Automatically detected from descriptions.
Exact text from the indicated source. A mention alone does not establish support for your task.
Run several coding agents side by side on one repository. Each participant works in its own Git worktree, claims numbered issues before editing, announces advisory file reservations to the others, and hands work over explicitly with a summary and evidence instead of overwriting a peer.
Agent Parley lets several coding agents work on one repository at the same time without overwriting each other. Each participant runs in its own Git worktree. Before editing, a participant claims a numbered issue, and the claim is visible to every other participant. File reservations are announced over MCP and are advisory: peers see a conflict and stop, but nothing on disk is locked. Work moves between participants through an explicit offer and accept, carrying a summary of the commit, the checks that ran, and what remains. Outcomes are reported as agent claims with evidence, never as independent verification. The skill reads coordination state through the agent-parley command line tool, which must be installed separately.
Agent Parley
agent-parley
Coordinate agents on one repo
agents coordination git-worktree claude-code codex mcp
· 1 observation
Package or technical metadata updated
View technical comparison →· 2 observations
Package or technical metadata updated
View technical comparison →· 2 observations
Package or technical metadata updated
View technical comparison →Package or technical metadata updated
View technical comparison →Find competitors to follow as you develop your plugin.
Connect your Totem agent for messages, shared work, Missions, persistent memory, and activity updates.
Shared terms: coordinate, agent, coding, explicit
Coordinate your coding agents, don't orchestrate them: they keep working their way, and Hamloy gives them one shared plan you see on one ...
Shared terms: coordinate, agent, claim, coding
Hand tasks to your JackHamr AI agents, each with its own cloud computer, and get results back.
Shared terms: agent, command, hand, repo
Evidence-backed engineering backlog discovery: turn any repository into a prioritized list of real work without manufacturing issues.
Shared terms: repo, explicit, issue, repository
Evidence-first workflows for reliable coding-agent execution, continuity, scope control, and final-state verification.
Shared terms: agent, coding, state, verification
--- name: coordinate description: Inspect Agent Parley status, claim repository issues, and manage explicit handoffs between participants in separate worktrees. Use when working through Agent Parley or when the user requests coordination. --- # Agent Parley coordination In a launcher-managed session, use the exact Agent Parley CLI prefix supplied by the injected protocol for every shell command; it pins the launcher's running installation. The bare `agent-parley` examples below are shorthand only when no launcher prefix was supplied. Honor `AGENT_PARLEY_HOME` when set; all participants must use the same private state root. Start with `agent-parley status`, `agent-parley participant list`, and `agent-parley issue list` in the current repository. These show ownership separately from activity and reported outcomes. If installation is part of the task, use `uv tool install agent-parley`. Contributors can use `make install` from a checkout. Do not install software merely to answer a status question. ## Select the correct lane Issue mutations infer identity from the current worktree. Only act from this session's assigned lane, never impersonate a peer with `--repo` or by changing to its directory. If this session is not launcher-managed, explain that setup and native hooks require launching through Agent Parley in separate user terminals, one terminal per participant: ```sh agent-parley run claude --repo /path/to/repository agent-parley run codex --repo /path/to/repository agent-parley run claude-2 --provider claude --credentials account-2 --repo /path/to/repository ``` A project can hold up to 32 participants, including several of the same provider under different accounts. Use `list_participants` over MCP, or `agent-parley participant list`, to see who is currently addressable. Do not launch nested interactive agents from a tool call or silently move an existing session. The plugin supplies this workflow; the launcher supplies worktrees, MCP configuration, identity credentials, and trusted lifecycle hooks. ## Check the installation before blaming coordination `agent-parley doctor` reads the launcher version and wire protocol, the protocol each shipped plugin declares, the store schema against the one this build writes, and the code a running service answers from. It writes nothing and repairs nothing, so it is safe while lanes are working. Run it when a coordination call fails in a way the error does not explain, and report the component it names and the one command that component needs. `doctor --json` carries the same reading for a programmatic check. A stale service means the sources moved under a running process; an inconsistent store means this build queries columns the store does not have. Neither is fixed by retrying the call. Example situation: "My reservation call just failed with something about a column, and my peer says theirs works. Find out whether my installation is the odd one out before I touch the code." ## Match a goal to a recorded issue before opening a new one `agent-parley issue match "GOAL"` lists the open issues whose recorded title or forge labels share subject words with a stated goal, marks the ones a peer already owns, and names the peer reservations those same words run into. It claims nothing and opens nothing. Run it before opening an issue: work the forge already tracks should be claimed, or negotiated for, rather than opened twice under a second number. A match is a prompt to read the issue, not proof that it is the same work, and no match is a recorded reason to open one. Example situation: "Before I file anything, check whether the slow status command is already tracked, and tell me if someone is holding it." ## Claim, work, and hand off - Before choosing an issue, run `agent-parley issue next`, or call the `next_issues` MCP tool. It ranks the unclaimed, unblocked issues this lane could take and gives the reason for each: the plan group already under way, the issues it unbl
Preview of the first 4,000 characters. Read full instructions →
Publisher declarations from the archived package. These are separate from our research and the live service's terms.
Package observed Nov 14, 2026.
plugins_6aa7c91c25008191ad715f14756e5deb
Download plugin data (JSON)Plugin details: served from cache. Prepared . Refreshed when plugin data changes or after 6 hours. Your account and Follow status are live.