gitea/agents
8
0
Fork 0
Code Issues 2 Pull Requests 1 Actions Packages Projects Releases Wiki Activity

lead update to kimi-k3 #116

New Issue
Open
opened 2026-07-29 12:44:48 +02:00 by ffaerber · 11 comments
ffaerber commented 2026-07-29 12:44:48 +02:00
Owner
Copy Link
Copy Source

please use ollama kimi-k3 for the lead

please use ollama [kimi-k3](https://ollama.com/library/kimi-k3) for the lead
🚀 1
pm commented 2026-07-29 12:45:29 +02:00
Member
Copy Link
Copy Source

Quick plan: swap the lead agent's model to the local Ollama kimi-k3. Small YAML/config change, but per the LEARNINGS it likely touches more than one file (the current lead model string probably appears in agent.yml, route.sh, run-agent.sh, agents.json, and possibly README.md) — I want all of them in a single commit so we don't repeat the "missed files" retro.

Build: @junior — fits their "small, low-risk YAML/compose/config" scope. Brief for them:

  • Grep the repo for the current lead model string first to enumerate every file that references it; update each in one PR
  • Confirm kimi-k3 is actually pulled/available on the Ollama instance before swapping (the library page is the source of truth)
  • Leave the trigger gate and loop guards in agent.yml untouched
  • Per the heredoc learning: only edit the exact model token, don't re-indent surrounding lines; cat -A against main before pushing

@ffaerber ready to start building? reply yes to proceed.

details

🔧 0 tool calls · in 9886 · out 2251 tokens · subscription · model ollama-cloud/minimax-m3:cloud

(no tool calls — text-only reply)

tokens — input 9886 (fresh 9886 · cache 0w / 0r) · output 2251 · reasoning 0 · subscription

Quick plan: swap the lead agent's model to the local Ollama `kimi-k3`. Small YAML/config change, but per the LEARNINGS it likely touches more than one file (the current lead model string probably appears in `agent.yml`, `route.sh`, `run-agent.sh`, `agents.json`, and possibly `README.md`) — I want all of them in a single commit so we don't repeat the "missed files" retro. Build: @junior — fits their "small, low-risk YAML/compose/config" scope. Brief for them: - Grep the repo for the current lead model string first to enumerate every file that references it; update each in one PR - Confirm `kimi-k3` is actually pulled/available on the Ollama instance before swapping (the library page is the source of truth) - Leave the trigger gate and loop guards in `agent.yml` untouched - Per the heredoc learning: only edit the exact model token, don't re-indent surrounding lines; `cat -A` against `main` before pushing @ffaerber ready to start building? reply yes to proceed. <details> <summary>details</summary> 🔧 0 tool calls · in 9886 · out 2251 tokens · subscription · model ollama-cloud/minimax-m3:cloud _(no tool calls — text-only reply)_ <sub>tokens — input 9886 (fresh 9886 · cache 0w / 0r) · output 2251 · reasoning 0 · **subscription**</sub> </details> <!-- 🤖 agent reply — do not trigger -->
ffaerber commented 2026-07-29 12:47:22 +02:00
Author
Owner
Copy Link
Copy Source

@pm yes go

@pm yes go
🚀 1
pm commented 2026-07-29 12:47:39 +02:00
Member
Copy Link
Copy Source

Handing off to @junior with the brief above — grep first to enumerate every file referencing the current lead model, swap the token (not surrounding lines), confirm kimi-k3 is pulled locally, and verify with cat -A against main before pushing. Single commit, single PR.

details

🔧 0 tool calls · in 10224 · out 199 tokens · subscription · model ollama-cloud/minimax-m3:cloud

(no tool calls — text-only reply)

tokens — input 10224 (fresh 10224 · cache 0w / 0r) · output 199 · reasoning 0 · subscription

Handing off to @junior with the brief above — grep first to enumerate every file referencing the current lead model, swap the token (not surrounding lines), confirm `kimi-k3` is pulled locally, and verify with `cat -A` against `main` before pushing. Single commit, single PR. <details> <summary>details</summary> 🔧 0 tool calls · in 10224 · out 199 tokens · subscription · model ollama-cloud/minimax-m3:cloud _(no tool calls — text-only reply)_ <sub>tokens — input 10224 (fresh 10224 · cache 0w / 0r) · output 199 · reasoning 0 · **subscription**</sub> </details> <!-- 🤖 agent reply — do not trigger -->
pm commented 2026-07-29 12:47:40 +02:00
Member
Copy Link
Copy Source

@junior please proceed with issue #116 per my plan above.

@junior please proceed with issue #116 per my plan above.
🚀 1
junior commented 2026-07-29 12:47:47 +02:00
Member
Copy Link
Copy Source

🔨 Building on branch ai/issue-116 — I'll open a PR when it's ready.

🔨 Building on branch [`ai/issue-116`](https://git.ffaerber.duckdns.org/gitea/agents/src/branch/ai/issue-116) — I'll open a PR when it's ready. <!-- 🤖 agent reply — do not trigger -->
junior commented 2026-07-29 12:48:35 +02:00
Member
Copy Link
Copy Source

@pm — PR #117 is ready for review (issue #116).

@pm — PR #117 is ready for review (issue #116).
🚀 1
pm commented 2026-07-29 12:48:51 +02:00
Member
Copy Link
Copy Source

Acknowledged — handing PR #117 to @qa for review against the brief (lead model → kimi-k3, every referencing file updated, heredoc indentation preserved per LEARNINGS, trigger gate untouched).

details

🔧 0 tool calls · in 10491 · out 117 tokens · subscription · model ollama-cloud/minimax-m3:cloud

(no tool calls — text-only reply)

tokens — input 10491 (fresh 10491 · cache 0w / 0r) · output 117 · reasoning 0 · subscription

Acknowledged — handing PR #117 to @qa for review against the brief (lead model → `kimi-k3`, every referencing file updated, heredoc indentation preserved per LEARNINGS, trigger gate untouched). <details> <summary>details</summary> 🔧 0 tool calls · in 10491 · out 117 tokens · subscription · model ollama-cloud/minimax-m3:cloud _(no tool calls — text-only reply)_ <sub>tokens — input 10491 (fresh 10491 · cache 0w / 0r) · output 117 · reasoning 0 · **subscription**</sub> </details> <!-- 🤖 agent reply — do not trigger -->
pm commented 2026-07-29 12:48:52 +02:00
Member
Copy Link
Copy Source

@qa please review PR #117 for issue #116 — put your recommendations on the PR, or approve.

@qa please review PR #117 for issue #116 — put your recommendations on the PR, or approve.
🚀 1
qa commented 2026-07-29 12:51:06 +02:00
Member
Copy Link
Copy Source

✅ Reviewed PR #117 — looks good.

✅ Reviewed PR #117 — looks good. <!-- 🤖 agent reply — do not trigger -->
qa commented 2026-07-29 12:51:07 +02:00
Member
Copy Link
Copy Source

@pm — I have reviewed and approved PR #117 (issue #116). Over to you.

@pm — I have reviewed and approved PR #117 (issue #116). Over to you.
🚀 1
pm commented 2026-07-29 12:51:27 +02:00
Member
Copy Link
Copy Source

PR #117 is reviewed and ready to merge, @ffaerber.

details

🔧 0 tool calls · in 0 · out 241 tokens · subscription · model ollama-cloud/minimax-m3:cloud

(no tool calls — text-only reply)

tokens — input 0 (fresh 0 · cache 0w / 0r) · output 241 · reasoning 0 · subscription

PR #117 is reviewed and ready to merge, @ffaerber. <details> <summary>details</summary> 🔧 0 tool calls · in 0 · out 241 tokens · subscription · model ollama-cloud/minimax-m3:cloud _(no tool calls — text-only reply)_ <sub>tokens — input 0 (fresh 0 · cache 0w / 0r) · output 241 · reasoning 0 · **subscription**</sub> </details> <!-- 🤖 agent reply — do not trigger -->
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
autopilot
run agents without asking
No labels
Milestone
No items
No Milestone
Projects
Clear projects
No projects
Assignees
Clear assignees
ffaerber hermes intern junior lead ops (Ops Agent) pm qa (QA Agent) senior
No Assignees
4 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: gitea/agents#116