docs(agents): what an agent must never invent - #180
Merged
Conversation
Two negative instructions, prompted by studying ctx's agent skill (their anti-hallucination clauses are the best part of that doc): when memesh is unavailable or recall returns nothing, say so and work without it — never fabricate a memory or cite a [mem:id] handle that was not actually returned; and recall windows are bounded, so a hit count is not a graph-wide count and an empty result is not proof of absence. A wrong 'remembered' fact arrives wearing the authority of the graph, which makes it worse than no memory at all.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Two negative instructions for agent-facing docs (AGENTS.md and the /memesh skill), prompted by studying ctx's agent skill — whose anti-hallucination clauses are the best-written part of that doc:
[mem:id]handle that was not actually returned. A wrong "remembered" fact is worse than no memory: it arrives wearing the authority of the graph.limit; a hit count is not a corpus count, and an empty result is not proof nothing was stored — vary the wording or narrow by tag before concluding.One wording lesson the doc-claims gate taught while writing this: lowercase
memesh <word>in agent docs is reserved for CLI-subcommand claims (the gate verifies each against registered commands), so product-name prose uses MeMesh.Verification
node scripts/run-tests-isolated.mjs→ exit 0, 142 files / 2126 tests.npm run verify:release→ exit 0 (incl. the agent-docs subcommand gate and the AGENTS.md tool-table parity gate).