loop-graph

Name: loop-graph
Author: levi-qiao

28stars2forks

How to install loop-graph?

levi-qiao/graphkit/loop-graph

$npx -y skills add levi-qiao/graphkit --skill loop-graph

Installs into the current project.

Summary

Run a long-horizon coding task as a small graph of agent nodes driven by two scheduled loops, instead of one drifting loop. Generates an executor node (works against a single-source-of-truth ledger) and a clean-context supervisor node that audits from outside the executor's conte

Files · 1

View on GitHub

SKILL.md

1	# loop-graph — a graph of agent nodes, not a drifting loop
2
3	## What it does & why
4
5	Turns a vague long-horizon request ("make it production-ready", "accuracy above baseline", "finish the migration") into a small graph of agent nodes that stays on-spec across many rounds:
6
7	- an executor node — drives the work round by round against a single scoreboard;
8	- a clean-context supervisor node — audits the work from outside the executor's context (re-verifying claimed-done against the shared standards) and corrects course before drift compounds.
9
10	Why a graph, not a loop: an agent grinding a long task is inside the context that drifted, so it rationalizes scope creep and calls half-done work "done". graphkit's nodes talk only through durable, inspectable state (ledger, git tree, directives file) — never a shared, polluted context. The supervisor boots fresh each tick and judges the run like an outside reviewer.
11
12	## The graph it builds
13
14	```
15	┌─────────────┐ reads / rewrites ┌──────────────┐
16	│ EXECUTOR │ ───────────────────▶ │ │
17	│ node │ ◀─────────────────── │ ledger.md │ ← single scoreboard
18	│ (does work) │ │ (shared state)│
19	└─────────────┘ └──────────────┘
20	▲ ▲
21	│ reads each round │ reads only (never writes it)
22	│ │
23	┌───────────────┐ appends corrections ┌──────────────────┐
24	│ directives.md │ ◀──────────────────── │ SUPERVISOR │ ← fresh, CLEAN context
25	└───────────────┘ │ node (watches) │ every tick
26	└──────────────────┘
27	│ checkpoint-commits clean work
28	│ escalates human-only calls
29	▼
30	git / you
31	```
32
33	Five artifacts from a short interview:
34
35	1. `executor.md` — executor prompt: one task book, one cadence, anti-bloat rules, stop conditions, red lines.
36	2. `ledger.md` — the single scoreboard both nodes read; the executor rewrites it every round.
37	3. `directives.md` — the one-way corrections edge, seeded near-empty (STANDING locks + numbered corrections the supervisor appends).
38	4. `ops.md` (optional) — durable env facts (build commands, credential/data policy) the executor consults, not re-derives.
39	5. `supervisor.md` (optional) — supervisor prompt, scheduled; independently re-verifies claimed-done work, checkpoint-commits what passes, decides pending items, and corrects drift / steers the plan via the directives file.
40
41	Run directory — fixed, one per run. Everything above goes in `.graphkit/<YYYY-MM-DD-slug>/` at the repo root (workspace root for multi-repo runs), plus an `archive/` subdir for in-run rotations. A new run gets a new directory, generated fresh from the templates — never retarget or edit a previous run's files: patching stale prompts wastes tokens and leaves leftover text steering toward the old goal. The old directory stays untouched (it is the archive); distill what still holds into the new ledger's starting snapshot and copy still-in-force STANDING directives forward. Commit the run directory unless data policy forbids — it's the durable state the graph depends on.
42
43	Invariants: ledger = the only scoreboard; one item per round → verify → update ledger; the supervisor steers only through `directives.md` (a one-way edge) — it never edits the ledger or shares the executor's context.
44
45	## When to use it
46
47	- Task spans many rounds; the user won't babysit each one.
48	- "Done" is verifiable (tests, gates, metrics) — the graph needs a definition of done it can check itself.
49	- Real risk of scope creep, "looks done" work, or quietly changing contracts / lowering the bar.
50
51	Not for a one-shot edit, or when success needs a human to judge every time — say so, suggest a

People who liked this also viewed

6 picks

Type

loop-graph

28stars2forks

View on GitHub

How to install loop-graph?

levi-qiao/graphkit/loop-graph

$npx -y skills add levi-qiao/graphkit --skill loop-graph

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# loop-graph — a graph of agent nodes, not a drifting loop
2
3	## What it does & why
4
5	Turns a vague long-horizon request ("make it production-ready", "accuracy above baseline", "finish the migration") into a small graph of agent nodes that stays on-spec across many rounds:
6
7	- an executor node — drives the work round by round against a single scoreboard;
8	- a clean-context supervisor node — audits the work from outside the executor's context (re-verifying claimed-done against the shared standards) and corrects course before drift compounds.
9
10	Why a graph, not a loop: an agent grinding a long task is inside the context that drifted, so it rationalizes scope creep and calls half-done work "done". graphkit's nodes talk only through durable, inspectable state (ledger, git tree, directives file) — never a shared, polluted context. The supervisor boots fresh each tick and judges the run like an outside reviewer.
11
12	## The graph it builds
13
14	```
15	┌─────────────┐ reads / rewrites ┌──────────────┐
16	│ EXECUTOR │ ───────────────────▶ │ │
17	│ node │ ◀─────────────────── │ ledger.md │ ← single scoreboard
18	│ (does work) │ │ (shared state)│
19	└─────────────┘ └──────────────┘
20	▲ ▲
21	│ reads each round │ reads only (never writes it)
22	│ │
23	┌───────────────┐ appends corrections ┌──────────────────┐
24	│ directives.md │ ◀──────────────────── │ SUPERVISOR │ ← fresh, CLEAN context
25	└───────────────┘ │ node (watches) │ every tick
26	└──────────────────┘
27	│ checkpoint-commits clean work
28	│ escalates human-only calls
29	▼
30	git / you
31	```
32
33	Five artifacts from a short interview:
34
35	1. `executor.md` — executor prompt: one task book, one cadence, anti-bloat rules, stop conditions, red lines.
36	2. `ledger.md` — the single scoreboard both nodes read; the executor rewrites it every round.
37	3. `directives.md` — the one-way corrections edge, seeded near-empty (STANDING locks + numbered corrections the supervisor appends).
38	4. `ops.md` (optional) — durable env facts (build commands, credential/data policy) the executor consults, not re-derives.
39	5. `supervisor.md` (optional) — supervisor prompt, scheduled; independently re-verifies claimed-done work, checkpoint-commits what passes, decides pending items, and corrects drift / steers the plan via the directives file.
40
41	Run directory — fixed, one per run. Everything above goes in `.graphkit/<YYYY-MM-DD-slug>/` at the repo root (workspace root for multi-repo runs), plus an `archive/` subdir for in-run rotations. A new run gets a new directory, generated fresh from the templates — never retarget or edit a previous run's files: patching stale prompts wastes tokens and leaves leftover text steering toward the old goal. The old directory stays untouched (it is the archive); distill what still holds into the new ledger's starting snapshot and copy still-in-force STANDING directives forward. Commit the run directory unless data policy forbids — it's the durable state the graph depends on.
42
43	Invariants: ledger = the only scoreboard; one item per round → verify → update ledger; the supervisor steers only through `directives.md` (a one-way edge) — it never edits the ledger or shares the executor's context.
44
45	## When to use it
46
47	- Task spans many rounds; the user won't babysit each one.
48	- "Done" is verifiable (tests, gates, metrics) — the graph needs a definition of done it can check itself.
49	- Real risk of scope creep, "looks done" work, or quietly changing contracts / lowering the bar.
50
51	Not for a one-shot edit, or when success needs a human to judge every time — say so, suggest a

People who liked this also viewed

6 picks

Type