Remove unsubstantiated performance metrics from architecture guide - #1
Merged
Merged
Conversation
The Performance Characteristics table presented invented success-rate figures (~80% first-pass, ~95% retry) as expected measurements. Drop the table until real usage data exists to back it. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014jSpHa2KzTZGTF1g9QsSPu
The planning skill instructed the Planner to address every risk the Explorer flagged, and gave concerns no home other than the plan itself — so models accommodated every potential issue up front and produced over-complicated plans. - Add a "Noted, not planned" carry-forward list to writing-plans: a first-class, one-line-per-item home for concerns that do not block the current milestone, reconsidered at phase boundaries - Replace "address each risk" with triage: plan only blocking risks, carry the rest forward - Add "Plan for Everything" anti-pattern (KISS/YAGNI) alongside "One Big Task" - Reframe phases as milestones with one concrete deliverable each; prefer more, smaller phases - Bias complexity classification downward when torn between classes - Have the Explorer tag each risk blocking or carry-forward so the Planner gets pre-sorted input - Persist carried-forward items via Backlog (draft tasks or parent description) so deferral never loses information - Reinforce the small-plan bias in the planner agent stub Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014jSpHa2KzTZGTF1g9QsSPu
./install.sh --global installs the agents, commands, skills, and plugins into ~/.config/opencode (respecting XDG_CONFIG_HOME); --project <path> installs into a project's .opencode directory. Re-running updates a previous install item by item, replacing what this repo provides while leaving the user's own agents, commands, and skills untouched. --symlink links items instead of copying so a git pull updates every install. Destination directories use the plural names from the current OpenCode docs (agents/, commands/). Repo-specific configuration (opencode.json, tui.json) is never installed. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014jSpHa2KzTZGTF1g9QsSPu
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The Performance Characteristics table presented invented success-rate
figures (~80% first-pass, ~95% retry) as expected measurements. Drop
the table until real usage data exists to back it.
Co-Authored-By: Claude Fable 5 noreply@anthropic.com
Claude-Session: https://claude.ai/code/session_014jSpHa2KzTZGTF1g9QsSPu