flagon-io/g1t

Pick any line to see why it is the way it is: the commit, the pull request and issue it came from, and what the agent was thinking.

Issues and pull requests replace intents and attempts10d ago		1	-- Issues, pull requests, comments and sessions.
		2	-- Every timestamp is RFC 3339 UTC text.
		3
		4	-- Issues and pull requests share one sequence of numbers per repository.
		5	CREATE TABLE counters (
		6	repo_id TEXT PRIMARY KEY,
		7	last INTEGER NOT NULL
		8	);
		9
		10	CREATE TABLE issues (
		11	id TEXT PRIMARY KEY,
		12	repo_id TEXT NOT NULL,
		13	number INTEGER NOT NULL,
		14	title TEXT NOT NULL,
		15	body TEXT NOT NULL,
		16	-- JSON array of label names.
		17	labels TEXT NOT NULL DEFAULT '[]',
		18	-- JSON array of commands.
		19	checks TEXT NOT NULL DEFAULT '[]',
		20	state TEXT NOT NULL DEFAULT 'open',
		21	-- Why it was closed: completed or not_planned.
		22	reason TEXT,
		23	-- The number of the pull request whose merge closed it.
		24	resolved_by INTEGER,
		25	author_id TEXT NOT NULL,
		26	author_name TEXT NOT NULL,
		27	created_at TEXT NOT NULL,
		28	updated_at TEXT NOT NULL,
		29	closed_at TEXT,
		30	UNIQUE (repo_id, number)
		31	);
		32	CREATE INDEX issues_by_state ON issues (repo_id, state, number);
		33
		34	CREATE TABLE pulls (
		35	id TEXT PRIMARY KEY,
		36	repo_id TEXT NOT NULL,
		37	number INTEGER NOT NULL,
		38	-- The issue it is for, if any, and that issue's number.
		39	issue_id TEXT REFERENCES issues (id),
		40	issue_number INTEGER,
		41	title TEXT NOT NULL,
		42	body TEXT,
		43	agent TEXT NOT NULL,
		44	runtime TEXT NOT NULL,
		45	status TEXT NOT NULL DEFAULT 'draft',
Pull requests from branches10d ago		46	-- Where the change is: a fork made for the pull request, or a branch of
		47	-- the repository itself.
		48	fork_repo_id TEXT UNIQUE,
		49	fork_namespace TEXT,
		50	fork_name TEXT,
		51	source_branch TEXT,
Issues and pull requests replace intents and attempts10d ago		52	head_commit TEXT,
		53	-- What the branch pointed to before a merged pull request landed.
		54	merge_base TEXT,
		55	merged_by TEXT,
		56	merged_at TEXT,
		57	-- The pull request merged instead of this one.
		58	superseded_by INTEGER,
Agents as a team: lifecycle, merge queue, billing and a new shell9d ago		59	-- JSON array of { path, additions, deletions }: what it changes, as of
		60	-- its latest push. NULL until first worked out.
		61	files TEXT,
Acceptance checks in sandboxes, line comments and review verdicts10d ago		62	-- The latest run of the issue's acceptance checks, and where it stands.
		63	check_run_id TEXT,
		64	check_status TEXT,
Issues and pull requests replace intents and attempts10d ago		65	author_id TEXT NOT NULL,
		66	author_name TEXT NOT NULL,
		67	created_at TEXT NOT NULL,
		68	updated_at TEXT NOT NULL,
		69	UNIQUE (repo_id, number)
		70	);
		71	CREATE INDEX pulls_by_status ON pulls (repo_id, status, number);
		72	CREATE INDEX pulls_by_issue ON pulls (issue_id);
		73	CREATE INDEX pulls_by_author ON pulls (author_id, status);
		74
		75	-- On an issue or a pull request: the two share numbers.
		76	CREATE TABLE comments (
		77	id TEXT PRIMARY KEY,
		78	repo_id TEXT NOT NULL,
		79	number INTEGER NOT NULL,
		80	author_id TEXT NOT NULL,
		81	author_name TEXT NOT NULL,
		82	body TEXT NOT NULL,
Acceptance checks in sandboxes, line comments and review verdicts10d ago		83	-- For a comment on one line of a pull request's change: the file, and
		84	-- the line as numbered after the change.
		85	path TEXT,
		86	line INTEGER,
		87	-- A reviewer's decision: approve or request_changes.
		88	verdict TEXT,
Issues and pull requests replace intents and attempts10d ago		89	created_at TEXT NOT NULL
		90	);
		91	CREATE INDEX comments_by_subject ON comments (repo_id, number, id);
		92
		93	CREATE TABLE session_entries (
		94	pull_id TEXT NOT NULL REFERENCES pulls (id),
		95	seq INTEGER NOT NULL,
		96	kind TEXT NOT NULL,
		97	text TEXT NOT NULL,
		98	tool TEXT,
		99	-- The fork's head when the entry was recorded: links reasoning to code.
		100	"commit" TEXT,
		101	at TEXT NOT NULL,
		102	PRIMARY KEY (pull_id, seq)
		103	);
Acceptance checks in sandboxes, line comments and review verdicts10d ago		104
		105	-- Runs of an issue's acceptance checks against a pull request's head.
		106	CREATE TABLE check_runs (
		107	id TEXT PRIMARY KEY,
		108	pull_id TEXT NOT NULL REFERENCES pulls (id),
		109	head_commit TEXT NOT NULL,
		110	-- queued, running, passed, failed or errored.
		111	status TEXT NOT NULL DEFAULT 'queued',
		112	-- JSON array of { command, passed, exitCode, output, durationMs }.
		113	results TEXT NOT NULL DEFAULT '[]',
		114	error TEXT,
		115	-- SHA-256 of the token the sandbox reports with.
		116	token_hash TEXT NOT NULL,
		117	created_at TEXT NOT NULL,
		118	finished_at TEXT
		119	);
		120	CREATE INDEX check_runs_by_pull ON check_runs (pull_id, id);
Agents as a team: lifecycle, merge queue, billing and a new shell9d ago		121
		122	-- Reviews of a pull request being written by a g1t agent in a sandbox.
		123	CREATE TABLE review_runs (
		124	id TEXT PRIMARY KEY,
		125	pull_id TEXT NOT NULL REFERENCES pulls (id),
		126	-- SHA-256 of the token the sandbox reports with.
		127	token_hash TEXT NOT NULL,
		128	created_at TEXT NOT NULL,
		129	finished_at TEXT
		130	);
		131	CREATE INDEX review_runs_by_pull ON review_runs (pull_id, id);

This file's history is long; its oldest lines are credited to the oldest commit read.