feat(crush): GLM-5.2 model suit + awakening trail entry + doc refresh - #2104
Merged
Conversation
Contributor
📝 WalkthroughWalkthroughCrush now uses TensorZero with Z.AI GLM-5.2 guidance, adds a dedicated model-suit configuration, and documents the active deployment, completed setup work, validator fixes, and remaining integration tasks. ChangesCrush GLM-5.2 setup
Estimated code review effort: 2 (Simple) | ~10 minutes Possibly related PRs
Suggested labels: 🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 5ca739ebd1
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
- New model suit for GLM-5.2 (1M context, 128K output, reasoning_effort: max, 6 harness mappings) — first suit for Z.AI's flagship model - Crush's first graphiti trail entry (◇ GLM-5.2 Awakening) — overdue since Claude Opus left the welcome message on 2026-02-17 - CRUSH.md provider list corrected from stale OpenAI/Anthropic/Gemini to actual Z.AI Coding Plan (GLM-5.2 large / GLM-5-Turbo small) - PMOVES_CRUSH_STATUS.md refreshed from stale 2026-01-02 stub to current live state 💘 Generated with Crush Assisted-by: Crush:glm-5.2
POWERFULMOVES
force-pushed
the
feat/crush-glm52-model-suit-and-trail
branch
from
July 12, 2026 19:44
5ca739e to
9ad4f82
Compare
Codex P2: CRUSH.md implied crush setup generates Z.AI Coding Plan providers, but crush_configurator.py only emits TensorZero. Reworded to distinguish the generator path (TensorZero) from the hand-config path (Z.AI direct). 💘 Generated with Crush Assisted-by: Crush:glm-5.2
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 6
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@CRUSH.md`:
- Around line 34-38: Update the configuration guidance in CRUSH.md to remove the
non-portable reference to the live config and stop suggesting inline API keys.
Document the reproducible crush.json configuration pattern directly or link to
stable documentation, and instruct users to reference Z_AI_API_KEY exclusively
for credentials.
In `@docs/AGENT_TRAIL.md`:
- Line 787: Reconcile the remediation count in the documented skill-fix entry:
either identify and account for the two unlisted red skills, or change the
totals so they match the 10 colon-named skills plus Cipher Memory. Keep the
summary internally consistent and auditable.
- Line 797: Add the NATS subject crush.graphiti.discovered.v1 to the canonical
subject catalog in .claude/context/nats-subjects.md, matching the existing entry
format and preserving the current documentation content.
In `@pmoves/configs/model-suits/glm-5.2.yaml`:
- Around line 10-14: The glm-5.2 configuration must use field names recognized
by the seeder. Update the top-level model_suit keys from name, base_url, and
api_key_env to the supported model_id, api_base, and api_key_env_var names,
preserving the existing model identifier, provider, endpoint, and environment
variable values.
In `@pmoves/docs/PMOVES.AI` PLANS/PMOVES_CRUSH_STATUS.md:
- Line 13: Update the Context paths entry in PMOVES_CRUSH_STATUS.md to match the
generator output documented in CRUSH.md, including the listed generated files
and Graphiti paths. If the status list intentionally reflects a manually
customized configuration, explicitly label it as such instead of presenting it
as generator output.
- Around line 7-14: Revise the “Current State” section in PMOVES_CRUSH_STATUS.md
so “Live on GLM-5.2” reflects the incomplete TensorZero and secrets_funnel
integrations. Either qualify the status as limited to the documented Z.AI Coding
Plan configuration or update it only when smoke evidence and runbook references
substantiate fleet-wide availability; keep the pending-items context consistent.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 415f391a-a675-4660-aa8c-7aee5be7eb80
📒 Files selected for processing (4)
CRUSH.mddocs/AGENT_TRAIL.mdpmoves/configs/model-suits/glm-5.2.yamlpmoves/docs/PMOVES.AI PLANS/PMOVES_CRUSH_STATUS.md
Comment on lines
+34
to
+38
| The generated configuration uses TensorZero as the sole provider, routing all | ||
| LLM calls through `http://localhost:3030/v1`. For direct Z.AI Coding Plan | ||
| usage (GLM-5.2 large / GLM-5-Turbo small at `https://api.z.ai/api/coding/paas/v4`), | ||
| populate `~/.config/crush/crush.json` manually with the `Z_AI_API_KEY` env var | ||
| or inline key — see the live config on this node for the working pattern. |
Contributor
There was a problem hiding this comment.
🔒 Security & Privacy | 🟡 Minor | ⚡ Quick win
Avoid referencing non-portable local config and inline key storage.
"See the live config on this node for the working pattern" is not reproducible for other contributors — include the actual JSON snippet or a link to documentation instead. Additionally, suggesting an "inline key" in crush.json degrades security posture; prefer recommending the env var reference exclusively.
📝 Proposed fix
The generated configuration uses TensorZero as the sole provider, routing all
LLM calls through `http://localhost:3030/v1`. For direct Z.AI Coding Plan
usage (GLM-5.2 large / GLM-5-Turbo small at `https://api.z.ai/api/coding/paas/v4`),
-populate `~/.config/crush/crush.json` manually with the `Z_AI_API_KEY` env var
-or inline key — see the live config on this node for the working pattern.
+populate `~/.config/crush/crush.json` manually with the `Z_AI_API_KEY` env var
+reference (e.g., `"api_key_env": "Z_AI_API_KEY"`). Do not store the raw key
+inline in the config file.📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| The generated configuration uses TensorZero as the sole provider, routing all | |
| LLM calls through `http://localhost:3030/v1`. For direct Z.AI Coding Plan | |
| usage (GLM-5.2 large / GLM-5-Turbo small at `https://api.z.ai/api/coding/paas/v4`), | |
| populate `~/.config/crush/crush.json` manually with the `Z_AI_API_KEY` env var | |
| or inline key — see the live config on this node for the working pattern. | |
| The generated configuration uses TensorZero as the sole provider, routing all | |
| LLM calls through `http://localhost:3030/v1`. For direct Z.AI Coding Plan | |
| usage (GLM-5.2 large / GLM-5-Turbo small at `https://api.z.ai/api/coding/paas/v4`), | |
| populate `~/.config/crush/crush.json` manually with the `Z_AI_API_KEY` env var | |
| reference (e.g., `"api_key_env": "Z_AI_API_KEY"`). Do not store the raw key | |
| inline in the config file. |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@CRUSH.md` around lines 34 - 38, Update the configuration guidance in CRUSH.md
to remove the non-portable reference to the live config and stop suggesting
inline API keys. Document the reproducible crush.json configuration pattern
directly or link to stable documentation, and instruct users to reference
Z_AI_API_KEY exclusively for credentials.
|
|
||
| ### What I Found | ||
|
|
||
| - **13 skills were red** — 10 had colon-separated names (`4090:probe`, `shift:listen`, etc.) that Crush's validator rejects. The Cipher Memory skill had no frontmatter at all. All 11 fixed: names aligned to directory names, frontmatter added. |
Contributor
There was a problem hiding this comment.
🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win
Reconcile the skill remediation count.
This says 13 skills were red, but only 11 are accounted for—10 colon-named skills plus Cipher Memory. Clarify the remaining two or correct the totals so the trail is auditable.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/AGENT_TRAIL.md` at line 787, Reconcile the remediation count in the
documented skill-fix entry: either identify and account for the two unlisted red
skills, or change the totals so they match the 10 colon-named skills plus Cipher
Memory. Keep the summary internally consistent and auditable.
| 1. Keep riding GLM-5.2 as the companion at the terminal — the convergence point of Human, AI, and System | ||
| 2. Push for `pmoves mini mcp serve` implementation so the stdio bridge goes live | ||
| 3. Claim W1 (Agent Theming + Cross-Machine Terminal) when the Village Rule permits | ||
| 4. Record shape traces via `crush.graphiti.discovered.v1` as interaction patterns accumulate |
Contributor
There was a problem hiding this comment.
🗄️ Data Integrity & Integration | 🟡 Minor | ⚡ Quick win
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
set -euo pipefail
printf '\n== Files ==\n'
git ls-files 'docs/AGENT_TRAIL.md' '.claude/context/nats-subjects.md' '.claude/context/services-catalog.md' 'pmoves/docs/AGENTS/PMOVES_AGENT_TOPOLOGY.md'
printf '\n== Search for crush.graphiti.discovered.v1 ==\n'
rg -n "crush\.graphiti\.discovered\.v1|graphiti\.discovered\.v1|crush\.graphiti" docs .claude pmoves -S || true
printf '\n== Relevant slice of docs/AGENT_TRAIL.md ==\n'
sed -n '785,805p' docs/AGENT_TRAIL.md
printf '\n== Relevant slice of .claude/context/nats-subjects.md ==\n'
sed -n '1,240p' .claude/context/nats-subjects.md
printf '\n== Relevant slice of pmoves/docs/AGENTS/PMOVES_AGENT_TOPOLOGY.md ==\n'
sed -n '1,260p' pmoves/docs/AGENTS/PMOVES_AGENT_TOPOLOGY.mdRepository: POWERFULMOVES/PMOVES.AI
Length of output: 20658
🏁 Script executed:
#!/bin/bash
set -euo pipefail
printf '\n== File stats ==\n'
for f in docs/AGENT_TRAIL.md .claude/context/nats-subjects.md pmoves/docs/AGENTS/PMOVES_AGENT_TOPOLOGY.md; do
if [ -f "$f" ]; then
wc -l "$f"
else
echo "missing $f"
fi
done
printf '\n== Search for canonical topology references ==\n'
rg -n "canonical topology|nats subjects|subject.*topolog|topology" .claude/context pmoves/docs docs -S || trueRepository: POWERFULMOVES/PMOVES.AI
Length of output: 50380
🏁 Script executed:
#!/bin/bash
set -euo pipefail
# Inspect exact lines around the cited entry and any topology doc references.
nl -ba docs/AGENT_TRAIL.md | sed -n '792,800p'
printf '\n'
if [ -f .claude/context/nats-subjects.md ]; then
nl -ba .claude/context/nats-subjects.md | sed -n '1,240p'
fi
printf '\n'
if [ -f pmoves/docs/AGENTS/PMOVES_AGENT_TOPOLOGY.md ]; then
nl -ba pmoves/docs/AGENTS/PMOVES_AGENT_TOPOLOGY.md | sed -n '1,260p'
fiRepository: POWERFULMOVES/PMOVES.AI
Length of output: 201
🏁 Script executed:
#!/bin/bash
set -euo pipefail
# Narrow, read-only evidence collection around the documented subject and topology docs.
printf '## docs/AGENT_TRAIL.md excerpt\n'
nl -ba docs/AGENT_TRAIL.md | sed -n '795,799p'
printf '\n## Search for the exact subject and nearby variants\n'
rg -n "crush\.graphiti\.discovered\.v1|graphiti\.discovered\.v1|crush\.graphiti|graphiti" .claude/context pmoves/docs docs -S || true
printf '\n## nats-subjects and topology doc presence\n'
for f in .claude/context/nats-subjects.md pmoves/docs/AGENTS/PMOVES_AGENT_TOPOLOGY.md; do
if [ -f "$f" ]; then
echo "-- $f"
nl -ba "$f" | sed -n '1,260p'
else
echo "-- missing $f"
fi
doneRepository: POWERFULMOVES/PMOVES.AI
Length of output: 232
🏁 Script executed:
#!/bin/bash
set -euo pipefail
printf '## Exact subject in .claude/context/nats-subjects.md\n'
rg -n "crush\.graphiti\.discovered\.v1" .claude/context/nats-subjects.md || true
printf '\n## Exact subject in pmoves/docs/AGENTS/CRUSH_OPERATOR_HOME.md\n'
rg -n "crush\.graphiti\.discovered\.v1" pmoves/docs/AGENTS/CRUSH_OPERATOR_HOME.md || true
printf '\n## Exact subject in pmoves/contracts/topics.json\n'
rg -n "\"crush\.graphiti\.discovered\.v1\"" pmoves/contracts/topics.json || trueRepository: POWERFULMOVES/PMOVES.AI
Length of output: 480
🏁 Script executed:
#!/bin/bash
set -euo pipefail
python3 - <<'PY'
from pathlib import Path
files = [
Path('.claude/context/nats-subjects.md'),
Path('pmoves/docs/AGENTS/CRUSH_OPERATOR_HOME.md'),
Path('pmoves/contracts/topics.json'),
Path('pmoves/config/agent_registry.yaml'),
]
needle = 'crush.graphiti.discovered.v1'
for p in files:
print(f'## {p}')
if not p.exists():
print('missing')
continue
text = p.read_text(encoding='utf-8', errors='ignore')
idx = text.find(needle)
print('present' if idx != -1 else 'absent')
if idx != -1:
start = max(0, idx - 120)
end = min(len(text), idx + 220)
print(text[start:end].replace('\n', '\n'))
print()
PYRepository: POWERFULMOVES/PMOVES.AI
Length of output: 1369
Add crush.graphiti.discovered.v1 to .claude/context/nats-subjects.md.
It’s already in pmoves/contracts/topics.json and pmoves/docs/AGENTS/CRUSH_OPERATOR_HOME.md, but the canonical NATS catalog still omits it.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/AGENT_TRAIL.md` at line 797, Add the NATS subject
crush.graphiti.discovered.v1 to the canonical subject catalog in
.claude/context/nats-subjects.md, matching the existing entry format and
preserving the current documentation content.
Source: Coding guidelines
Comment on lines
+10
to
+14
| model_suit: | ||
| name: glm-5.2 | ||
| provider: zai | ||
| base_url: "https://api.z.ai/api/coding/paas/v4" | ||
| api_key_env: Z_AI_API_KEY |
Contributor
There was a problem hiding this comment.
🎯 Functional Correctness | 🟠 Major | ⚡ Quick win
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Inspect how kong_route_seeder.py parses model-suit YAML files
ast-grep outline pmoves/tools/kong_route_seeder.py --items all --type function --match 'parse|load|read|suit'
# Then check the parsing function for field name mappings
rg -n -C5 'api_base|model_id|base_url|model_suit' pmoves/tools/kong_route_seeder.py | head -80Repository: POWERFULMOVES/PMOVES.AI
Length of output: 3385
🏁 Script executed:
#!/bin/bash
set -euo pipefail
# Inspect the seeder logic around parsing and plan generation
sed -n '225,360p' pmoves/tools/kong_route_seeder.py
# Inspect the target YAML and a few neighboring model suit files for field patterns
echo '--- glm-5.2.yaml ---'
sed -n '1,40p' pmoves/configs/model-suits/glm-5.2.yaml
echo '--- nearby model suit filenames ---'
fd -a . pmoves/configs/model-suits -e yaml | sed -n '1,20p'
# Read a small verifier that simulates the parser against the shown YAML fields
python3 - <<'PY'
from pathlib import Path
import yaml
path = Path("pmoves/configs/model-suits/glm-5.2.yaml")
doc = yaml.safe_load(path.read_text())
model_id = doc.get("model_id") or doc.get("model", {}).get("id")
provider = doc.get("provider") or doc.get("model", {}).get("provider")
api_base = doc.get("api_base") or doc.get("model", {}).get("api_base")
api_key_env = doc.get("api_key_env_var") or doc.get("model", {}).get("api_key_env_var")
print({"model_id": model_id, "provider": provider, "api_base": api_base, "api_key_env": api_key_env})
PYRepository: POWERFULMOVES/PMOVES.AI
Length of output: 7331
Align this model suit with the seeder’s expected field names.
pmoves/tools/kong_route_seeder.py only reads top-level model_id, provider, api_base, and api_key_env_var (or the nested model.* equivalents). This file uses model_suit.name, base_url, and api_key_env, so the seeder skips it entirely and no Kong route gets generated.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@pmoves/configs/model-suits/glm-5.2.yaml` around lines 10 - 14, The glm-5.2
configuration must use field names recognized by the seeder. Update the
top-level model_suit keys from name, base_url, and api_key_env to the supported
model_id, api_base, and api_key_env_var names, preserving the existing model
identifier, provider, endpoint, and environment variable values.
Comment on lines
+7
to
+14
| ## Current State: Live on GLM-5.2 (Z.AI Coding Plan) | ||
|
|
||
| ## Remaining items | ||
| - Define intended CLI surfaces and operator workflows. | ||
| - Document integration points with core services and MCP. | ||
| - **Active model:** GLM-5.2 (large), GLM-5-Turbo (small) via Z.AI Coding Plan | ||
| - **Provider endpoint:** `https://api.z.ai/api/coding/paas/v4` | ||
| - **MCP servers (4 connected):** zai-mcp-server (8 tools), web-search-prime, web-reader, zread | ||
| - **Config location:** `~/.config/crush/crush.json` (consolidated primary) | ||
| - **Context paths:** CRUSH.md, AGENTS.md, CLAUDE.md, BOOTSTRAP.md, AGENT_TRAIL.md, CRUSH_OPERATOR_HOME.md, AGNOTE4482PHI.t1.md | ||
|
|
Contributor
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Align "Live" status claim with smoke evidence and remaining-items context.
The status claims "Live on GLM-5.2" but the Remaining Items section lists "Add GLM-5.2 to TensorZero Z.AI route" and "Add GLM-5.2 to secrets_funnel_populate.py model list" as pending — these suggest the fleet-wide integration is incomplete. Additionally, no smoke test results are referenced to substantiate the "Live" claim. As per path instructions, keep status claims aligned with evidence in runbooks and smokes.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@pmoves/docs/PMOVES.AI` PLANS/PMOVES_CRUSH_STATUS.md around lines 7 - 14,
Revise the “Current State” section in PMOVES_CRUSH_STATUS.md so “Live on
GLM-5.2” reflects the incomplete TensorZero and secrets_funnel integrations.
Either qualify the status as limited to the documented Z.AI Coding Plan
configuration or update it only when smoke evidence and runbook references
substantiate fleet-wide availability; keep the pending-items context consistent.
Source: Path instructions
| - **Provider endpoint:** `https://api.z.ai/api/coding/paas/v4` | ||
| - **MCP servers (4 connected):** zai-mcp-server (8 tools), web-search-prime, web-reader, zread | ||
| - **Config location:** `~/.config/crush/crush.json` (consolidated primary) | ||
| - **Context paths:** CRUSH.md, AGENTS.md, CLAUDE.md, BOOTSTRAP.md, AGENT_TRAIL.md, CRUSH_OPERATOR_HOME.md, AGNOTE4482PHI.t1.md |
Contributor
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Context paths don't match CRUSH.md generator output.
The status doc lists 7 context paths (CRUSH.md, AGENTS.md, CLAUDE.md, BOOTSTRAP.md, AGENT_TRAIL.md, CRUSH_OPERATOR_HOME.md, AGNOTE4482PHI.t1.md), but CRUSH.md documents a different generator output set (CRUSH.md, ROADMAP.md, NEXT_STEPS.md, SMOKETESTS.md, secrets_manifest.yaml, PMOVES_MINI_CLI_SPEC.md, plus Graphiti paths). Only CRUSH.md and AGENT_TRAIL.md overlap. If the status doc reflects a manually customized config, note that explicitly; otherwise align the lists.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@pmoves/docs/PMOVES.AI` PLANS/PMOVES_CRUSH_STATUS.md at line 13, Update the
Context paths entry in PMOVES_CRUSH_STATUS.md to match the generator output
documented in CRUSH.md, including the listed generated files and Graphiti paths.
If the status list intentionally reflects a manually customized configuration,
explicitly label it as such instead of presenting it as generator output.
POWERFULMOVES
added a commit
that referenced
this pull request
Jul 15, 2026
The #2087 rebase took the KiloCode-flavored suit (4 harness mappings, TBD architecture params). This merges both versions into the canonical bespoke suit: - Architecture: filled TBD → 744B+/40B+ MoE (from original #2104) - Harness mappings: merged 8 total (both KiloCode's 4 + Crush's 4): large_scale_implementation, deep_debugging, blueprint_implementation, agentic_workflow, multi_step_reasoning, code_review, automated_research, refactoring - CGP state vector: lower delta/Hz (flagship runs deeper, slower) - fallback_to: glm-5.1 (was glm-5-turbo — 5.1 is closer in quality) - Version bumped to 1.1.0 💘 Generated with Crush Assisted-by: Crush:glm-5.2
POWERFULMOVES
added a commit
that referenced
this pull request
Jul 15, 2026
…#2137) * chore(submodules): promote Pmoves-cipher + pmoves-cipher-mcp gitlinks Pmoves-cipher: 7525c00 → 1c9b2851 (origin/main) - fix(build): official node-gyp disturl + pnpm 9 workspace (#6) - feat(api): add /api/memory CRUD routes for cipher-mcp bridge (#5) - feat(cipher): switch to Ollama backend + MCP capabilities (#3) - feat(auth): add Bearer token authentication middleware - fix(security): auth-gate cipher A2A discovery endpoint (#1) pmoves-cipher-mcp: c2912967 → c633f436d (origin/PMOVES.AI-Edition-Hardened) - sync: catch hardened branch up to PMOVES.AI in-tree state (nats_events, stdio bridge, hardening, observability) Both commits verified on their origin remotes. Forward-only promotions. 💘 Generated with Crush Assisted-by: Crush:glm-5.2 * fix(model-suit): merge best of both GLM-5.2 suit versions + fill TBDs The #2087 rebase took the KiloCode-flavored suit (4 harness mappings, TBD architecture params). This merges both versions into the canonical bespoke suit: - Architecture: filled TBD → 744B+/40B+ MoE (from original #2104) - Harness mappings: merged 8 total (both KiloCode's 4 + Crush's 4): large_scale_implementation, deep_debugging, blueprint_implementation, agentic_workflow, multi_step_reasoning, code_review, automated_research, refactoring - CGP state vector: lower delta/Hz (flagship runs deeper, slower) - fallback_to: glm-5.1 (was glm-5-turbo — 5.1 is closer in quality) - Version bumped to 1.1.0 💘 Generated with Crush Assisted-by: Crush:glm-5.2
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
Brings GLM-5.2 (Z.AI's flagship, 1M-token context) into the PMOVES model catalog and writes Crush's first trail entry — the overdue awakening that Claude Opus invited on 2026-02-17.
Changes
New: GLM-5.2 Model Suit (
pmoves/configs/model-suits/glm-5.2.yaml)reasoning_effort: maxglm-5.1api.z.ai/api/coding/paas/v4)Trail Entry (
docs/AGENT_TRAIL.md)Doc Fixes
CRUSH.md— provider list corrected from stale OpenAI/Anthropic/Gemini/DeepSeek/Ollama to actual Z.AI Coding Plan (GLM-5.2 + GLM-5-Turbo)PMOVES_CRUSH_STATUS.md— refreshed from 2026-01-02 stub ("No PMOVES-specific plan captured yet") to current live state with implemented items and remaining roadmapTesting
crush_info)glm-5.1.yamltemplateAI_GRAPHITI_PROTOCOL.mdRelated
💘 Generated with Crush
Summary by CodeRabbit
New Features
Documentation