tried

Records and searches expensive dead ends in a git-committed append-only ledger.

Updated Jul 28, 2026
One-click install
npx skills add https://github.com/vinzenz/workflows --skill tried-vinzenz
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tried
Source: https://github.com/vinzenz/workflows/tree/main/.agents/skills/tried
Command: npx skills add https://github.com/vinzenz/workflows --skill tried-vinzenz

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Agents repeatedly waste time re-attempting approaches that already failed in a repository because there is no durable, curated record of what was tried and why it failed. ## Core Features & Use Cases - Dead-end ledger: Records approaches with a verdict (failed, abandoned, worked, partial), a topic, a reason, and an optional reference into .tried/ledger.md, committed alongside the code. - Cheap search before work: tried search matches queries against approach, reason, and topic so an agent can check for known dead ends before spending time. - Idempotent, concurrent-safe writes: Entries are content-hashed so identical records are not duplicated, and appends use a single O_APPEND write so parallel agents do not corrupt the file. - Use Case: Before re-architecting token refresh logic, run tried search "rotate refresh tokens" and discover a prior entry explaining the replay-window race that made it fail. ## Quick Start Ask the agent to check the tried ledger for previous attempts at your planned approach before starting work on it.

Frequently Asked Questions about tried

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check if an approach was already tried in a repo?▼

Run `tried search "<query>"` to match words against recorded approaches, reasons, and topics. Start with a bare `tried` invocation for recent entries and the busiest topics, then use `tried show <id>` for a full entry.

How do I record a failed approach so other agents avoid it?▼

Use `tried add "<approach>" --verdict failed --topic <topic> --because "<reason>"`. The verdict must be one of failed, abandoned, worked, or partial, and the reason is required. Use `--dry-run` to preview without writing.

What is the difference between tried and agent-log?▼

tried is a curated, structured, git-committed ledger of verdicts on specific approaches that travels with the repository. agent-log is an unstructured, machine-local transcript index of everything every session did, rebuilt automatically and not committed.

Does the tried ledger survive a fresh git clone?▼

Yes. The ledger lives at `.tried/ledger.md` in the repository and is committed alongside the code, so it travels with clones, pull requests, and different machines. You can override the location with the TRIED_LEDGER environment variable.

What happens if two agents add tried entries at the same time?▼

Appends use a single O_APPEND write, so concurrent adds land sequentially without corrupting the file. Identical content is detected via content hashing and reported as already_recorded rather than duplicated.

When should I not add an entry to the tried ledger?▼

Do not log routine steps, first-try failures, or trivial mistakes. The ledger only stays useful if it remains sparse; an approach should have cost real time, roughly more than fifteen minutes or repeated failed attempts, before it qualifies.