0 Fork 0

mirror of https://github.com/ryan4yin/nix-config.git synced 2026-11-13 12:13:46 +01:00

Files

T

ryan4yin | 二花 373b233f32 feat(agents): steerable turns, PTC, and a Nushell-only dsh (#473 )

* feat(agents): use the shell's own dialect; allow 5 s retry backoff

Assisted-by: DeepSeek Harness (deepseek-flash)

* feat(dsh): make Nushell the only shell

Assisted-by: DeepSeek Harness (deepseek-flash)

* docs(agents): wait for CI in the background

Pushes to a PR often start multi-minute CI runs, and an agent that holds
the turn open on a foreground watch reports nothing until it ends. Require
the pushed head and running checks to be reported first, the watch to run
in the background, and the result to be collected after independent work
or after ending the turn.

Assisted-by: DeepSeek Harness (deepseek-flash)

* docs(agents): keep the shell-handoff rule platform-neutral

The global rules carry no Windows assumptions: drop the cmd, pwsh, and
wsl handoff examples and keep the POSIX forms plus the dsh refusal.

Assisted-by: DeepSeek Harness (deepseek-flash)

* docs(dsh): keep the shell-swap docs POSIX-only

This repository manages Linux and Darwin hosts, not Windows: drop the
Windows handoff examples and terminal-backend wording from dsh.md. The home
patch also disabled pwsh-sandbox redundantly because dsh-base already
disables that row on POSIX; keep only the active bash-sandbox disable and
verify the composed rows with --dump-config.

Assisted-by: DeepSeek Harness (deepseek-flash)

* docs(agents): keep the global rules runtime-neutral

Global rules are loaded by every agent CLI, so they must not describe one
runtime's behavior. Drop the dsh-specific sentence and wording while keeping
the expectation to prefer Nushell structured pipelines and Python. Runtime
behavior stays documented in home/base/tui/agents/dsh.md.

Assisted-by: DeepSeek Harness (deepseek-flash)

* docs(agents): default shell work to Nushell

Keep the Nushell preference explicit in the runtime-neutral global rules, and
keep runtime behavior in runtime docs. Bash stays a narrow fallback for tasks
Nushell cannot do.

Assisted-by: DeepSeek Harness (deepseek-flash)

* docs(agents): tighten global-rule scope and CI guidance

Keep global rules runtime- and project-neutral, and state that boundary in the
file. Remove the duplicated CI turn-steering text in favour of the Communication
rule, resolve the watch run before watching instead of relying on
gh pr checks --watch alone, name the working-directory parameter generically,
and distinguish a bounded retry backoff from a fixed completion wait.

Assisted-by: DeepSeek Harness (deepseek-flash)

* docs(agents): never estimate a CI run's duration

A CI run has no predictable length, so do not turn "it looks short" into a
reason to wait. The rule is unconditional: report the head and started checks,
start a background watch, and end the turn; collect the result on completion.
The eval scenario now uses an unknown duration.

Assisted-by: DeepSeek Harness (deepseek-flash)

* docs(agents): move maintenance guidance to the README

The runtime/project boundary is guidance for editing the rules, not a rule
agents need loaded every session. The README's Maintaining global rules
section already owns that audience; the loaded rules keep only behavior.

Assisted-by: DeepSeek Harness (deepseek-flash)

* refactor(dsh): render the home patch from a YAML template

Move the home-layer cordis patch into dsh-cordis.patch.yml, substitute the
bundle and Nushell store paths with replaceVars, and link the rendered file
through home.file.source. This module is the only writer of
~/.dsh/cordis.patch.yml, the current host has no such file, and the profile
patch is a separate layer.

Assisted-by: DeepSeek Harness (deepseek-flash)

* refactor(dsh): install the shell plugin through NUR withPlugins

Replace the local runCommand bundle, peer symlink, and home-layer YAML file
with the plugin package and dsh.withPlugins from ryan4yin/nur-packages#43. The
package wrapper adds a generated --patch overlay after the profile layer, so
dsh no longer writes or manages ~/.dsh/cordis.patch.yml. The NUR input is
temporarily pinned to the PR branch and its commit; point it back at main once
the NUR PR merges.

Assisted-by: DeepSeek Harness (deepseek-flash)

* refactor(nix): expose the NUR package set as nurPkgs

Pass one per-system NUR package set through specialArgs instead of repeating
nur-ryan4yin.packages.${pkgs.stdenv.hostPlatform.system} at each call site.
The agents, rime, and idols-ruby modules now use the short binding.

Assisted-by: DeepSeek Harness (deepseek-flash)

* refactor(nix): use nurPkgs in package overlays

The overlays receive specialArgs too, so they can use the per-system NUR
package set instead of rebuilding the long input path from prev. This leaves
only the definition in outputs/default.nix.

Assisted-by: DeepSeek Harness (deepseek-flash)

* refactor(nix): name the per-system NUR set myNurPkgs

Leave room for consuming another NUR repository later, and match the existing
mylib/myvars naming.

Assisted-by: DeepSeek Harness (deepseek-flash)

* chore(dsh): consume the merged NUR package set

nur-ryan4yin now points at main after ryan4yin/nur-packages#43 was squash
merged; the lock records the merge commit. No branch ref remains.

Assisted-by: DeepSeek Harness (deepseek-flash)

* fix(dsh): pass the web profile explicitly

The withPlugins launcher prepends --patch, which historically prevented dsh
from auto-expanding "dsh web" into "--profile web". Pass the profile
explicitly in the Just recipe and the verification commands so they work with
both plain dsh and the wrapper.

Assisted-by: DeepSeek Harness (deepseek-flash)

* chore(dsh): pick up the wrapped-profile fix

Update nur-ryan4yin to the squash merge that keeps an implicit profile name
first, so "dsh web" and "just dsh-web" work through the wrapper again.

Assisted-by: DeepSeek Harness (deepseek-flash)

* chore(dsh): pick up the plugin peer-link fix

Update nur-ryan4yin to the squash merge that links the dsh peer tree at
node_modules/@deepseek-ai instead of nesting it one level deeper. This is the
import failure that stopped nushell-executor from activating after deploy.

Assisted-by: DeepSeek Harness (deepseek-flash)

2026-10-18 12:28:18 +08:00

8.0 KiB

Raw Permalink Blame History

Agents

Portable agent resources shared across projects: the global baseline rules, the global skills, behavioral scenarios, and reference snippets for external tooling.

Home Manager links the rules and skills into the agents' config locations. The resource files here remain portable Markdown; deployment lives in the Home Manager module.

Repo-scoped task procedures for one repository belong elsewhere: put them in that repository's .agents/skills/ (note the leading dot), which agents discover automatically. In a repository, use the layers this way:

AGENTS.md: always-loaded map and hard safety rules; keep it short.
agents/skills/*/SKILL.md: custom global procedures reused across repositories.
*.md / README.md: reference and domain runbooks for people, kept next to the code they describe.
.agents/skills/*/SKILL.md: procedures specific to the current repository.

Keep one canonical home for each fact; link between layers instead of copying paragraphs.

What this directory contains

global-rules.md: global baseline rules for coding agents. It is deliberately not named AGENTS.md: agents working in this checkout would load it a second time as nested instructions for agents/.
skills/: global skills kept in this repository, e.g. git-delivery for commit messages, PR descriptions, review replies, and cleanup after merge; its README records every global skill and its source.
evals/global-rules.md: how to size and run evaluations for changes to the rules or skills, and the behavioral scenarios.
install-tooling.md: install snippets for project-scoped skills and other external agent tooling.

The Nix side of the agents — deploying the rules and skills, the agent CLIs, and their environment — lives in the Home Manager module home/base/tui/agents/.

Core workflow

Maintain shared boundaries in agents/global-rules.md and task procedures in skills.
Configure permissions directly in the agent runtime; auto-approval is generally used.
Content changes to the rules and to skills in agents/skills/ reach the next agent session through out-of-store links; a running harness may need a reload. Adding or removing a skill, or moving a skill flake input, requires a Home Manager switch, run by the user.
Use install-tooling.md as a reference when installing project-scoped tooling.

Maintaining global rules

Before adding or expanding a rule, read the current rules and identify the decision boundary that needs to change. Keep reusable boundaries in global-rules.md; put concrete incidents, bypass attempts, and counterexamples in behavioral scenarios. Keep the file runtime- and project-neutral: runtime behavior belongs in that runtime's documentation, project mechanics in the project's instructions, and a rule belongs here only when it applies across both.

If an existing rule already covers the incident, add or refine a scenario instead of another rule.
If the boundary is missing or ambiguous, amend the relevant rule rather than append a special case.
Keep authorization, trust, and secret-handling boundaries always loaded. Put command examples and task-specific procedures in reference docs or skills.
Review the net growth and remove repetition. Brevity must preserve the boundary; verify both the prohibited action and the authorized action still behave as intended.

After changing the rules or a custom skill, size the evaluation to the change as the evaluation guide describes, and record the results in the PR.

Deployment

rules.nix links global-rules.md into every supported agent config directory as AGENTS.md, out-of-store, so edits apply without a rebuild. The per-agent target list lives in that module's README.

skills.nix links every directory under agents/skills/ and the skills selected from pinned flake inputs into ~/.agents/skills/<name>; see the module's global skills section. To add a local skill, create agents/skills/<name>/SKILL.md; the user activates it with a Home Manager switch. Links point at the canonical checkout, so changes in another worktree deploy only after they are merged there.

The module is imported through home/base/tui, so it covers the hosts that import home/linux/gui.nix or the macOS home/darwin stack; core-only servers are unchanged.

Permission configuration is not managed by Nix. The repository-root AGENTS.md contains guidance for this Nix configuration repository only; it is not deployed.

Auto-approval controls tool prompting. The global rules still define task authorization, safety, and secret handling.

TODO

Remaining work on the agent context, ordered by expected payoff. The goal is fewer tokens per session and closer adherence to these rules. Done so far: the rules source rename, the move of global skills to flake inputs (dropping superpowers, find-skills, and caveman), the i-have-adhd always-on block, find-docs rewritten onto the nixpkgs ctx7 CLI, the scenario and skill tables converted to lists, and the duplication cleanup in the repo skills. Historical behavioral runs cover tool-execution compression, whole-file compression, and boundaries for readiness waits, process inspection, and shell command sequencing on gpt-6-luna, the local model, and gpt-6.1-sol. A later full-scenario-set paper pass ran on DeepSeek-V41-Flash, with focused verification on that model and gpt-6-luna. The added boundaries cost more than the compression saved; revision-specific sizes, results, and coverage limits are recorded in #428.

Fold mattpocock/skills pr into git-delivery: before/after evidence, one-way or two-way door, and blast radius for PR bodies.
Write a testing-strategy skill: E2E of real workflows first, unit tests for logic and edge cases E2E cannot reach.
Merge ponytail-review into the shipped review-comments-and-docs skill (its doc-pruning half came from MidAutumnMoon/TaysiTsuki home/agents/skills/prune-comments-and-docs). Worth borrowing from garrytan/gstack review/checklist.md: a critical pass before an informational one, reading consumers outside the diff for a new enum value or loosened validator, a do-not-flag list, and "zero findings is valid". Keep review report-only.
Adapt wayfinder in the fork so its map and tickets default to local Markdown and research stays local, instead of GitHub issues and pushed research/ branches.
Adapt domain-modeling in the fork to record decisions in the repository's existing documents (here AGENTS.md, WORKAROUNDS.md, SECURITY.md, READMEs) and create GLOSSARY.md or ADRs only where none exist.
Borrow small gstack ideas into the rules: completion statuses (done, done with concerns, blocked, needs context); never call a failure pre-existing without running the same check on the base branch; a subagent with no one to ask takes the recommended option but never a destructive or publishing one.
Use progressive disclosure for large repo skills. nix-config-umu-game (~3.5k tokens) keeps its troubleshooting inline; move it to a reference file. Lower priority: nix-config-secrets §7 and nixpkgs-review §3.
Budget the always-loaded context in CI: cap skill description length and rules size, as gstack does for its catalog.
MCP servers are still configured by hand per harness; they could come from the same Nix source, though config formats differ per client. Reference: mirkolenz/infra options/home-manager/agents.nix.

Conventions

Keep files portable and reviewable.
Keep secrets and machine-specific credentials out of this directory.
Keep guidance generic enough to reuse across multiple agent environments.

8.0 KiB Raw Permalink Blame History