cd /news/ai-agents/show-hn-git-worktrees-and-evidence-g… Β· home β€Ί topics β€Ί ai-agents β€Ί article
[ARTICLE Β· art-34269] src=github.com β†— pub= topic=ai-agents verified=true sentiment=↑ positive

Show HN: Git worktrees and evidence gates for Codex and Claude Code

Developer Alex Reysa released glueRun-go, an open-source orchestration engine that runs autonomous AI coding agents in parallel against software repositories using a three-tier scheduling model with git-worktree isolation. The tool supports Claude and Codex agents, implementing durable leases, state packets, and gate/audit pipelines to manage multi-agent coding workflows. It aims to improve developer productivity by enabling safe parallel AI-driven code generation and testing.

read8 min views1 publishedJun 19, 2026
Show HN: Git worktrees and evidence gates for Codex and Claude Code
Image: source
      _          ___
 __ _| |_  _ ___| _ \_  _ _ _ ___ __ _ ___
/ _` | | || / -_)   / || | ' \___/ _` / _ \
\__, |_|\_,_\___|_|_\\_,_|_||_|  \__, \___/
|___/                            |___/

Autonomous multi-agent orchestration for software repos. One engine, many consumers.

glueRun-go is a bash + Python orchestration engine that drives autonomous AI coding agents in parallel against a repository. It implements a three-tier scheduling model (L0 origin loop β†’ L1 area planners β†’ L2 worker agents) with durable leases, state packets, gate/audit pipelines, and git-worktree isolation. The engine is installed once per machine and pinned per consumer repo β€” improvements propagate by bumping a version pin, not by re-copying scripts.

Tier Role
L0 origin
The single scheduler. Runs the reconcile cycle: import β†’ recover β†’ integrate β†’ dispatch β†’ snapshot. Holds the origin lock only during control work.
L1 area planners
One planner per DAG node (area). Reads the node's context, plans a batch of L2 tasks, and stages them as proposals for L0 to import.
L2 workers
Execute a single task in an isolated git worktree on a per-task branch. Produce a state packet (owned files, changes, evidence). An auditor reviews the packet; the decider routes the outcome.

Each gluerun reconcile --actuate

runs:

Importβ€” pull staged L1 task proposals into the DAG under the origin lock.** Recover**β€” reclaim stale leases whose workers have exited or timed out.** Integrate**β€” merge completed worker branches into the target branch under the git-op lock.** Dispatch**β€” pre-lease frontier tasks and spawn L2 workers.** Snapshot**β€” write a human-readable project state snapshot.

Every in-flight task holds a lease (a JSON file in .gluerun-state/leases/

) that records ownership, retry count, and expiry. When a worker finishes it writes a state packet (state-packet.v0.schema.json

) enumerating owned files, changed files, commands, tests, and evidence. The auditor validates the packet; the reaper attributes outcomes on later cycles.

After each L2 worker run the host executes the configured gate command (e.g. npm test

). A gate result (gate-result.v0.schema.json

) feeds the auditor model, which returns an audit verdict (audit-verdict.v0.schema.json

). The decider maps the (failure-class, retries-left)

pair to a recovery action β€” retry, amend-scope, escalate, or park β€” using a deterministic fast-path table before falling back to a model round-trip.

Detached dispatch is ON by default. When GLUERUN_DETACHED_DISPATCH=1

(the default), reconcile

pre-leases each frontier task and spawns the worker in its own session via dispatch-wrap.sh

, then returns within seconds. The origin lock is held only for the cycle's control work. A reaper (gluerun_reap_dispatches

) runs at the top of every apply/actuate cycle and attributes completions, failures, and crashes by checking dispatch records + worker exit files (pid liveness defeats pid reuse; crash detection drops from the 60-min stale-lease window to ~one cycle).

This is what keeps import, integrate, recover, STATUS, and STOP responsive while long workers run in the background.

Set GLUERUN_DETACHED_DISPATCH=0

to restore the legacy synchronous batch path, where reconcile

waits for every worker before returning.

Prerequisites:

  • Bash >= 4, python3

, andgit

. - At least one supported runner CLI on PATH

(claude

,codex

, or another configured runner). - macOS users may need brew install bash

and aPATH

entry that resolvesbash

to the Homebrew version before/bin/bash

.

git clone https://github.com/alex-reysa/glueRun-go /path/to/glueRun-go
cd /path/to/glueRun-go
bash install.sh

export PATH="$HOME/.gluerun/bin:$PATH"

In each consumer repo:

gluerun init      # scaffold gluerun.config.json, docs/orchestration/, .gluerun-version
gluerun doctor    # check deps, engine resolution, repo config

Each repo pins its engine version in .gluerun-version

(overrides gluerun.config.json

engineVersion

). The gluerun

launcher resolves that version from ~/.gluerun/versions/<ver>

, binds GLUERUN_ROOT

to the current repo, loads its config, and execs the engine. Run gluerun update <ver>

to repin.

gluerun reconcile --actuate

gluerun drive TASK-0001

gluerun auto

gluerun reconcile --drain

All per-repo variation lives in the consumer repo, never in engine files:

β€” declarative:gluerun.config.json

targetBranch

,gateCommand

,runner

,areas{}

,areaPrefix

,prewarm

,modules[]

,identity{}

,env{}

,provisionFiles[]

,envAllowlist[]

.β€” optional shell extras (computed values, functions).gluerun.config.sh

β€” gitignored operator overrides and secrets..gluerun-state/config.local.sh

The starter config deliberately sets gateCommand

to false

so a newly scaffolded repo fails closed until you replace it with the command that proves the repo is healthy.

provisionFiles

entries copy repo-local, gitignored files into each worker worktree after git worktree add

: { "source": ".env.local", "target": ".env.local", "required": true }

. The source and target must both be ignored or provisioning fails closed. envAllowlist

accepts exact env names or prefix patterns ending in *

; allowed values are written to worktree/.gluerun-state/worktree-env.sh

and sourced for prewarm/gate phases.

Env knob Default Effect
GLUERUN_MAX_CONCURRENT
5
Maximum L2 workers running concurrently.
GLUERUN_MAX_DISPATCH
5
Maximum tasks dispatched per reconcile cycle.
GLUERUN_DETACHED_DISPATCH
1
Default ON. Reconcile spawns workers in their own session and returns in seconds; the reaper attributes outcomes on later cycles. Set 0 for the legacy synchronous batch wait.
GLUERUN_AUTO_INTEGRATE
1
Automatically integrate (merge) completed worker branches in direct reconcile --actuate , gluerun auto , launchd, and console-driven cycles.
GLUERUN_PUSH
0 direct / 1 auto
Push integrated branches to the remote. Direct engine commands default local-only; gluerun auto /launchd set 1 unless overridden.
GLUERUN_MAX_HOURS
12
Wall-clock budget for the autonomy loop (gluerun auto ).
GLUERUN_MAX_RETRIES
3
Per-task worker retries before the decider escalates.
GLUERUN_STALE_MINUTES
60
Lease age (minutes) before a task without a live dispatch pid is reclaimed by the reaper.
GLUERUN_TARGET_BRANCH
(required)
Integration target branch in the consumer repo.
GLUERUN_SESSION_AFFINITY
1
Reuse a role's prior runtime session when all staleness gates pass; 0 always runs fresh.
GLUERUN_FIX_PROMPT_STRUCTURED
1
Structured fix prompt on retries (authoritative findings); 0 = legacy fix_hints tail.
GLUERUN_DECIDER_FAST
1
Resolve clear-cut failure classes by host policy table; 0 routes every failure through the model decider.
GLUERUN_WORKER_INFRA_MAX
1
Extra worker re-runs on an infra failure before surfacing worker-infra .
GLUERUN_AUDIT_INFRA_MAX
2
Extra auditor re-runs on an infra failure before surfacing audit-infra .
GLUERUN_CONTEXT_SECTION_MAX_CHARS
4000
Per-section cap on continuity content appended to prompts.
GLUERUN_PREFLIGHT_REQUIRE_ACCEPTANCE
1
Preflight requires non-empty acceptanceCriteria on a task.

Between retry attempts glueRun-go carries authoritative state forward rather than re-deriving it from a log tail:

Context capsulesβ€” hash-stampedimplementer-capsule.json

andreviewer-capsule.json

per attempt.Findings ledgerβ€”findings-status.json

upserted from each audit verdict, with stable finding ids tracked open/resolved across retries.Structured fix promptsβ€” the worker receives authoritative open findings on retry (setGLUERUN_FIX_PROMPT_STRUCTURED=0

to revert to the legacy byte-tail).Re-audit delta promptsβ€” the auditor receives prior findings + fix diff + per-id verification targets.** Attempt archive**β€” each attempt's artifacts are copied (never moved) underruns/<id>/attempts/<n>/

with anattempts/index.json

.

Optional role-keyed runtime session resume (codex exec resume

, claude -r

) behind 10 staleness gates, defaulting ON (GLUERUN_SESSION_AFFINITY=1

). Any gate failure or runner that refuses the resume degrades silently to a fresh run within the same attempt.

Invariant:session resume is a token-cost optimization that never changes a task outcome.

The generic engine/

references zero project-specific symbols β€” enforced by tests/test-engine-clean.sh

(the abstraction gate test). All per-project logic lives in opt-in modules:

gluerun-ext/
  storage-proof.sh    # example: durable-proof regime
  promote-gate.sh     # example: gate promoter

Modules are listed in gluerun.config.json

β†’ modules[]

. A repo that doesn't list them never loads them. The GLUERUN_MODULES

env var is the runtime list (set by the JSON config ).

Two versions move independently:

Engine pinβ€”.gluerun-version

is the canonical per-repo pin (overridesgluerun.config.json

engineVersion

; if they disagree.gluerun-version

wins andgluerun doctor

warns).gluerun update <ver>

rewrites it.Schemaβ€”SCHEMA_VERSION

(repo root) holds the data-contract version (v1

today). A repo records the schema it was scaffolded against ingluerun.config.json

β†’schemaVersion

.gluerun doctor

fails on a schema mismatch;gluerun migrate

runs the shippedmigrations/<from>-to-<to>.sh

chain and rewritesschemaVersion

. All runtime JSON schema identifiers follow the namespacegluerun.orchestration.*.v0

.

bash tests/run.sh    # 23 regression tests

The test suite uses no live state β€” all fixtures use a generic layer vocabulary. The tests/test-engine-clean.sh

gate enforces the abstraction contract on engine/

.

Run bash tests/run.sh

before opening a PR. Keep engine/

generic: project-specific rules belong in opt-in modules under gluerun-ext/

or in a consumer repo's config. Do not commit .gluerun-state/

, .worktrees/

, .gluerun-evidence/

, local env files, or generated run artifacts.

glueRun-go executes repo-configured shell commands and launches local coding agents in git worktrees. Review gluerun.config.json

, gluerun.config.sh

, and task files before running it in an untrusted repo. Report vulnerabilities through GitHub's private vulnerability reporting for this repository; if that is unavailable, open a minimal public issue asking for a private channel and do not include exploit details.

Licensed under GPL-3.0 β€” see LICENSE.

── more in #ai-agents 4 stories Β· sorted by recency
── more on @alex reysa 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain β€” perfect for shipping the agent you just read about.

$git push zahid main
β†’ Live at https://your-agent.zahid.host βœ“
Get free account β†’ Pricing
from €0/mo Β· no card required
LIVE [news/show-hn-git-worktree…] indexed:0 read:8min 2026-06-19 Β· β€”